AI News Archive: June 5, 2026 — Part 2
Sourced from 500+ daily AI sources, scored by relevance.
- How xAI Went From Chasing Anthropic to Powering It
How xAI Went From Chasing Anthropic to Powering It The Information
Score: 68🌐 MovesJun 5, 2026https://www.theinformation.com/articles/xai-went-chasing-anthropic-powering - Video AI Wars: How Chinese Labs Are Winning The Race OpenAI Abandoned
Five Chinese video AI models to watch: ByteDance’s Seedance, Alibaba’s Wan and Happy Horse, Kuaishou’s Kling, MiniMax’s Hailuo AI and Tencent’s Hunyuan.
- Apple's WWDC: Tim Cook's AI legacy at stake in his final developer conference as CEO
Apple heads into WWDC with Tim Cook’s AI legacy, Siri’s future as an agentic platform, and the stock’s rich valuation all on the line.
- Canada joins EU in push for tech sovereignty with new AI strategy
A new national AI strategy puts sovereignty front and centre as Canada moves to reduce its dependence on foreign cloud and AI providers. Read more: Canada joins EU in push for tech sovereignty with new AI strategy
Score: 68🌐 MovesJun 5, 2026https://www.siliconrepublic.com/business/canada-joins-eu-in-push-for-tech-sovereignty-with-new-ai-strategy - US lawmakers propose the ‘Great American AI Act’ to block state AI rules
The proposed 'Great American AI Act' would bar states from passing laws governing AI model development, including their testing before release, but the states can still regulate AI systems. The post US lawmakers propose the ‘Great American AI Act’ to block state AI rules appeared first on MEDIANAMA .
Score: 67🌐 MovesJun 5, 2026https://www.medianama.com/2026/06/223-us-bill-great-american-ai-act-state-ai-rules/ - ICE’s Plan to Let Cops Around the Country Scan Faces to Verify Immigration Status
ICE plans to give potentially more than a thousand agencies access to a facial recognition app that verifies a person's immigration status.
Score: 66🌐 MovesJun 5, 2026https://www.404media.co/ices-plan-to-let-cops-around-the-country-scan-faces-to-verify-immigration-status/ - Meta’s smart glasses might soon sport facial recognition — and the code to power this dystopian feature is already present in the Meta AI app on your phone
The Meta AI app has quietly had facial recognition code added to it, which could one day allow the company's smart glasses to identify people.
- Data Center Developer Switch in Talks to Raise Billions at $50 Billion-Plus Valuation
Data Center Developer Switch in Talks to Raise Billions at $50 Billion-Plus Valuation The Information
- Australia’s intelligence agency, ASD, scored access to Anthropic’s controversial Claude Mythos AI
Australia's intelligence agency gets elite access to Anthropic's powerful Claude Mythos AI to reshape cybersecurity.
- New data center routing design cuts AWS networking energy costs by 40%, Amazon claims
Amazon has started deploying a completely new routing architecture in AWS data centers which it says will deliver higher throughput from fewer physical switches while slashing electricity consumption. The company claims the architecture, dubbed Resilient Network Graphs (RNG) by the AWS Networking Lab researchers who developed it, offers a more efficient alternative to the traditional ‘fat tree’ topology that dominates data centers today. According to Amazon’s overview , RNG has been the default routing architecture for most new AWS data centers since April, spurred by the architecture’s ability to deliver 33% better throughput from 69% fewer routers. Importantly, in an industry where operating costs are always a focus, using fewer switch-routers has led to a projected reduction in network infrastructure electricity consumption of 40%. “For customers, it means more resilient infrastructure behind every API call, database query, and machine learning training job, without changing a single line of code,” said Amazon’s researchers. Random graph theory Tech is overwhelmed with big claims, especially regarding energy efficiency, which has turned out to be a fundamental limit in an era where power consumption is a major constraint. Does this one stand up? The answer to this question begins with the limitations of today’s fat tree routing. First used in 1990’s supercomputing, fat tree routing was widely adopted in the 2000s because it scaled well to handle the huge data center bandwidth demands. Fat tree is hierarchical: Switch-router infrastructure is layered, and packets move up and down these layers with the structure dictating how the packets find the shortest path. The drawback is that, as data center networks get bigger, the architecture requires ever more switch and cabling infrastructure to maintain throughput. In practice, this means that data center designers are forced to cut corners for cost reasons, leading to higher congestion. A theoretical alternative that’s been discussed for years is to use a non-hierarchical ‘random graph’ topology, for example, the one proposed by Jellyfish , a university project from 2012. In principle, this is more efficient; switches connect to each other randomly in a flat mesh that avoids the need for multiple switch-routing layers. It is also more fault tolerant, Amazon’s researchers explained: “No single router is more important than any other. The loss of 1% of routers results in a roughly 1% capacity loss.” Unfortunately, the random graph topology has downsides, principally the need for impossibly complex cabling between switches across varying distances inside a data center. It also requires each node to hold a huge routing table that sets out every possible data path in its memory. Quasi-random Amazon’s researchers say they solved this by developing a new routing algorithm called ‘Spraypoint’ which combines the basic idea of a random graph topology with some of the hierarchy of fat tree to effect a “quasi-random” compromise. Traffic is randomly ‘sprayed’ to neighbors, giving it a wide selection of possible paths to its destination. But as packets get near to their destination, they are routed via ‘waypoint’ switches using a conventional shortest path algorithm. However, the biggest innovation is a new type of data center device called a ‘ShuffleBox’. This concentrates the complex wiring normally required in random graph topologies into a single box, allowing random interconnection between switches without long cable runs. Although the efficiencies claimed for RNG have not been independently verified, the fact that Amazon plans to make the architecture its default for most new data centers offers some validation. “The first quasi-random network went live near Dublin, Ireland, at the end of 2024, carrying real production traffic. We validated performance against the mathematical predictions, identified operational refinements, and applied them in two additional deployments,” said Amazon. Not for everyone Ryan Ries , chief AI and data scientist at AWS consultancy and MSP Mission Cloud, was positive about the development. “Across the industry, there is growing pushback on data center expansion, tied to energy demand, water use, and local community impact, so power and water performance have become two of the most important issues facing cloud providers today,” said Ries. “The efficiency claims are credible because AWS is saying RNG is already in production, and it’s now the default architecture for most new builds globally.” An obvious upside with RNG, added Amruth Laxman of cloud VoIP provider 4Voice, is that it proves that random graph features can be built into data center networks after all. However, its proprietary nature means that its direct influence is likely to be limited for now. “AWS is known to design most of their networking equipment. The big question at this point is how flexible they have made their design. Most hyperscale customers are not able to absorb the costs, while AWS has the resources to absorb the entire redesign costs,” he pointed out. Re-equipping existing data centers with any radically new technology would incur significant expense, which is why Amazon only plans to use RNG in new data centers, he noted, so, in the immediate future, “don’t expect other companies to copy this design.”
- Industry coalition urges Trump administration to take urgent action as AI data centers' extreme memory consumption threatens other industries — AI-driven memory chip shortage could raise prices in automotive, medical, telecommunications sectors
A coalition of nine U.S. trade associations has urged the Trump administration to address an AI-driven memory chip shortage, warning that soaring DRAM prices and constrained supply could raise costs for consumer electronics, automobiles, medical devices, and broadband infrastructure while disrupting supply chains through at least 2027.
- Meta wants to turn a billion customer chats into enterprise AI agents
2026 is turning out to be a pivotal proving ground for agentic AI: providers now need to show proven use cases, not just prototypes. Meta says it can deliver on this with agents contextualized by billions of customer conversations. The company is expanding its Meta Business Agent and rolling out a new Meta Business Agent Platform to provide infrastructure to allow enterprises to build, customize, and deploy agents at scale. Experts point out that the company must truly prove its differentiation in a crowded space dominated by the likes of Salesforce, Google, Microsoft, ServiceNow, and others. Technology analyst Carmi Levy noted that it feels somewhat “off-brand” for Meta to be targeting the same large-enterprise workloads as those in which agentic AI platformers are already establishing early momentum. At the same time, he said, “Meta can’t afford to fall off the back of the enterprise AI agent trend, particularly in the kinds of platforms being frantically deployed to take over workflows currently handled by large conventionally human workforces.” Surfacing insights across messaging apps Meta says its new Business Agent can help enterprises not only increase their output, but deliver “more relevant, personalized experiences from the very first interaction” across WhatsApp, Messenger, and Instagram. The agent can plug directly into existing enterprise infrastructure and answer business-specific questions based on internal data, make product recommendations from business catalogs, close sales, and book appointments and qualify incoming leads, the company said. While they can work autonomously in the background, admins can also set thresholds beyond which they require human intervention. Enterprises can activate Business Agents now, and, while initial setup is free, Meta will roll out paid subscription offerings “with options for businesses of every size,” the company says. The agent will also be able to act as a business partner of sorts, providing morning briefings to catch users up on chats they missed overnight and provide insights on threads. It will soon be expanded to handle more complex tasks like surfacing product insights or performing market research. This feature is being rolled out to select customers on the WhatsApp Business app, Instagram Pro, Messenger, and Meta Business Suite; other potential customers can join a waitlist . For enterprises looking for a more hands-on approach, Meta is introducing a new Business Agent Platform. With it, enterprises can build, customize, and deploy AI agents , and set “enterprise-grade” controls, guardrails, measurements, and rules, the company said. They will also be able to connect to systems like Shopify, Zendesk, and Shopee, and give agents the ability to act autonomously. The new offering currently works alongside Meta’s WhatsApp Business Platform, with support for Messenger and Instagram. The platform is live now, and businesses will be charged based on AI token usage, according to a Meta spokesperson. This means customers will only pay for what the agent actually does, and Meta will provide guidance on average token usage per interaction so enterprises can plan their spend. Messages will be priced separately. Agent context based on billions of threads Meta isn’t starting from scratch with its agentic platform, the spokesperson pointed out; more than one million businesses are using its agents in one billion-plus threads across WhatsApp, Messenger, and Instagram to respond to customers on a 24/7 basis “AI is going to take this to the next level, so this context becomes intelligence for the agent, which it can then use to send personalized marketing messages, follow up on orders, and build loyalty,” the spokesperson said. Meta says it has the unique ability to deploy an agent natively into the apps that 3.5 billion people are already using today. The company can help enterprises connect with customers and engage via agents, and is “really well positioned to offer both reach and context from these existing conversations and engagement that no other platform has,” the spokesperson said. How Meta can differentiate itself in the agent space These broadened agentic capabilities come as Meta’s foundational social media networks move through “mid-life maturity,” Levy noted. The company must maximize growth and margins to fund its ambitious AI buildout, marking a “crucial inflection point for a company struggling for relevance in this fast-emerging digital economy.” Meta’s advantage lies in its presence within SMEs, many of whom have spent years building communities across its diverse consumer-facing platforms, Levy said. Adding agentic AI competencies to familiar interfaces and established networks certainly makes it an easier sell in that segment. “Beyond that scale, however, Meta faces an uphill battle against agentic vendors who have already built trust relationships with larger clients with more complex workflow needs,” he noted. Security glitches still an issue From a security standpoint , however, Meta still has some significant kinks to work out with its agents. For instance, Malwarebytes Labs reported that Meta’s customer service chatbots handed over access to (mostly dormant) Instagram accounts with little effort on the part of hackers; they simply used VPNs to match the target account’s location, then kicked off a normal password reset, prompting the chatbot to provide a one-time reset code via email. This was successful because the bot was what security researchers call a “confused deputy.” It had access to Meta’s account management system and permissions, but wasn’t taught how to verify whether it was interacting with the real account holder. The company said it has since resolved the problem and secured exposed accounts. That isn’t a positive for enterprises, Levy noted: “Meta’s checkered AI security track record doesn’t necessarily engender the kind of trust necessary for enterprise decision makers to deploy these tools at scale.”
- Microsoft AI chief says company was “set free” from OpenAI to pursue superintelligence
For three years, Microsoft's artificial intelligence story has been inseparable from OpenAI. The partnership — cemented by a cumulative investment exceeding $13 billion — gave Microsoft early access to the most advanced AI models on the planet, catapulting its Copilot products into the enterprise mainstream and adding hundreds of billions of dollars to its market capitalization. To the outside world, Microsoft's AI strategy was OpenAI. Mustafa Suleyman wants to change that narrative. In an exclusive sit-down interview with VentureBeat at Microsoft Build 2026 , the CEO of Microsoft AI disclosed that a contractual change with OpenAI roughly six months ago granted his division the formal authority to pursue what he openly calls "superintelligence" — using Microsoft's own researchers, its own data pipelines, and its own custom silicon. "We were only sort of set free from our contract with OpenAI about six months ago to formally pursue superintelligence," Suleyman said. "So this is very early days." The comment, delivered matter-of-factly backstage at the Fort Mason Center here, offers the clearest signal yet of a strategic inflection point unfolding inside the world's most valuable public company. Microsoft is not abandoning OpenAI. But it is building something alongside it — and, eventually, something that could stand entirely on its own. Microsoft's first in-house model family signals a new level of AI ambition The most tangible evidence of that shift arrived the same day. Microsoft announced a family of seven new AI models developed entirely in-house by its AI Superintelligence Team, spanning reasoning, code generation, image creation, transcription, and voice synthesis. The models — branded under the "MAI" family name — are Microsoft's most ambitious first-party AI release to date. The flagship, MAI-Thinking-1 , is a 35-billion-active-parameter reasoning model that Microsoft says matches leading models in its weight class on key software engineering benchmarks and demonstrates advanced mathematical reasoning. Suleyman emphasized one point repeatedly: the model was trained from scratch on clean, commercially licensed data, without distillation from third-party frontier models — a direct, if unstated, contrast to the widespread industry practice of using outputs from competitors' systems to train cheaper alternatives. "We train our reasoning models from scratch," Suleyman wrote in a blog post accompanying the announcement. "We don't distill from other labs and we don't rely on unlicensed or opaque data." The rest of the family fills out a multimodal portfolio designed for enterprise deployment: MAI-Code-1-Flash , a lightweight coding model built specifically for GitHub Copilot and VS Code ; MAI-Image-2.5 , which supports both text-to-image and image editing; MAI-Transcribe-1.5 , which Microsoft claims is the most accurate transcription model available, operating across 43 languages; and MAI-Voice-2 , a multilingual speech-generation system. All of the models ship through Microsoft Foundry , the company's model-hosting and deployment infrastructure, and for the first time, developers can tune model weights themselves through third-party platforms including OpenRouter , Fireworks , and Baseten . But Suleyman made clear in the interview that the seven models are a proof of concept, not a finished product. The real project is the lab itself. "Our job is to make sure that when we look out to 2030 and beyond, we have the capacity not just to buy models from third parties, but to build the absolute frontier, the best models in the world," he said. "That's a long transition." What "set free" from OpenAI actually means for Microsoft's AI future To understand what Suleyman means by "set free," you need to understand the unusual contractual architecture that has governed Microsoft's AI efforts for years. When Microsoft invested billions into OpenAI beginning in 2019, the partnership came with a specific arrangement: OpenAI would build the frontier models, and Microsoft would serve as the exclusive cloud provider , integrating those models into its products and reselling them through Azure. The deal gave Microsoft extraordinary commercial leverage — access to the world's most advanced AI without having to build it — but it also created a dependency. Microsoft was explicitly barred from pursuing its own AGI research, and the agreement even capped how large a model the company could train, restricting it from building systems beyond a certain computing threshold measured in FLOPS. That arrangement was formally renegotiated. As Fortune and Axios reported in November, a revised deal with OpenAI removed those restrictions, clearing the way for Suleyman to launch the MAI Superintelligence Team and pursue what he calls " humanist superintelligence ." The result, in Suleyman's telling at the time, was a "best-of-both environment, where we're free to pursue our own superintelligence and also work closely with them." By the time he sat down with VentureBeat at Build 2026, roughly six months had passed since that self-sufficiency effort formally began. Microsoft had already started shipping in-house models — including MAI-Image-2-Efficient , a lighter-weight image generation model released in April — but the seven MAI models announced at Build are the team's most ambitious release yet: a full multimodal family spanning reasoning, code, image generation, transcription, and voice. Even so, Suleyman does not view the shift as a rupture with OpenAI. He described Microsoft's current position as one of abundance, not scarcity. "There's no immediate urgent need to fill a gap in three months' time or six months' time," he said. "We have OpenAI, we have Anthropic, we have thousands of models inside Foundry. So there's already a huge amount of optionality available to us." The framing is telling. Microsoft's push into first-party frontier models is not born out of a crisis in the OpenAI relationship but out of a strategic calculation: as AI becomes the most consequential technology layer in enterprise computing, the company cannot afford to depend entirely on partners for the foundational capability. "Over the next five years, we have to be able to produce state-of-the-art frontier-scale models," Suleyman said. "That's our mission." Suleyman says the shift from chatbots to autonomous AI agents has already begun If the seven MAI models represent the technical ambition, a new capability called Frontier Tuning represents the commercial logic. Announced alongside the models at Build, Frontier Tuning allows enterprise customers to customize MAI models using their own proprietary data, workflows, and domain terminology, all within their own secure compliance boundary. The system uses reinforcement learning environments — what Microsoft calls " training gyms for AI " — that let agents learn directly from real workplace tasks without affecting production systems. The results Microsoft shared are striking. An MAI model tuned for Excel reportedly matches GPT 5.4 performance while operating at up to ten times greater efficiency. Early enterprise adopters are seeing similar gains: when tuned for one unnamed organization's exacting standards, the MAI model achieved the highest win rate of any model tested at roughly one-tenth the cost. Suleyman framed Frontier Tuning as part of a broader evolutionary stage — a move from intelligence to action. "We've basically moved beyond just conversation," he told VentureBeat. "Now we're moving to action." He introduced a new framework for thinking about that progression: the shift from IQ (factual intelligence) to EQ (emotional intelligence, or the ability to follow tone and style instructions) to what he calls AQ — the "Actions Quotient." Future AI agents, in Suleyman's telling, won't just answer questions. They will log into enterprise software, navigate complex multi-application workflows, and execute tasks across Excel, Word, Teams, Jira, Adobe InDesign, and customer relationship management systems — just as a human employee would. "You should be able to show up on day one and almost provision credentials to a new AI agent," he said. "The model needs to be able to move across all of these different environments, and that's actually the great strength of Microsoft." The Build 2026 announcements bore this out in concrete product terms. Microsoft Scout , the company's first "Autopilot" agent, operates as an always-on background assistant built on the open-source OpenClaw technology. It runs with its own governed identity inside Microsoft Entra , so its actions are auditable and attributable. Windows 365 for Agents gives AI agents their own managed Cloud PCs, allowing them to interact directly with applications and browsers inside enterprise environments. And the Foundry platform received major updates — including hosted agents with sub-100-millisecond cold starts, a new Microsoft Agent Framework, and one-click publishing to Teams and Microsoft 365 Copilot. Why Microsoft believes enterprise data is the next AI training frontier Suleyman also articulated why he believes Microsoft's position is uniquely defensible — and the argument has less to do with model architecture than with where work actually happens. "We've sort of hoovered up all of the obvious pools of training data," he said, referring to the industry's early scramble to ingest the open web. "In the next phase, we actually want to be able to give these agents to companies to train on their specific tasks with the data that they have inside of their own big workflows." The claim is subtle but consequential. The first wave of generative AI was trained on publicly available text — books, websites, Reddit posts, code repositories. That data is now largely exhausted, and its use is increasingly contested in court. The next wave, Suleyman argues, will be trained on enterprise-specific data: the internal workflows, decision traces, and institutional knowledge that define how real organizations operate. Microsoft, which serves 493 of the Fortune 500 through Azure according to Suleyman, is already embedded inside those workflows through Microsoft 365, Teams, Dynamics 365, and the broader Azure ecosystem. Frontier Tuning is the mechanism that converts that positional advantage into model performance. "People underappreciate that that's going to be the next domain," Suleyman said. The early partner list for Frontier Tuning reflects the ambition: Mayo Clinic , where Microsoft is co-creating a frontier AI model for healthcare using de-identified clinical data; EY , which is tuning a tax-advisory agent for deployment to 75,000 professionals globally; Land O'Lakes , where Frontier Tuning delivered what the company's product development scientist called "meaningful improvements in grounded outputs and style compliance"; and Pearson , which is using tuned models to provide learning-science-aligned feedback in its Communication Coach product. The Mayo Clinic partnership may be the most significant. Microsoft and Mayo Clinic are collaborating to build a healthcare-specific frontier model that combines Mayo's clinical expertise and longitudinal patient insights with Microsoft's AI capabilities. The model will be owned by Mayo Clinic and deployed first within Mayo's own environment before being made available to other organizations through Foundry. Microsoft's custom AI chips and GPU buying spree reveal the scale of its compute advantage None of this works without an industrial-scale compute infrastructure, and Suleyman was unusually candid about the hardware economics underlying Microsoft's strategy. "We are the largest buyer of GPUs on the planet," he said. "We're the largest buyer of GB200s and GB300s in the world." Microsoft will continue purchasing Nvidia accelerators "for many, many years to come," Suleyman said. But the company is simultaneously building its own custom silicon. Maia 200 , Microsoft's second-generation AI accelerator, is already running in production across data centers in Iowa and Arizona, with deployments planned for Italy, Australia, and South Korea. According to Microsoft, Maia 200 delivers the best tokens-per-dollar-per-watt in the company’s fleet. Suleyman put a finer point on the economics in the interview: Maia 200 is 30 percent more cost-efficient than Nvidia's GB200, he said. And when Microsoft co-optimizes its own MAI models to run natively on Maia silicon, the company sees an additional 1.4x improvement in performance per watt. "It is going to be cheaper in years to come to build on MAI models with Maia 200 and Maia 300 inside of Azure," he said. That claim — if it holds at scale — has profound implications for the competitive landscape. It means Microsoft is not merely buying its way to AI dominance through Nvidia; it is building a vertically integrated stack in which its own models, running on its own chips, inside its own cloud, tuned on its customers' own data, could offer performance and cost characteristics that no competitor can replicate. Suleyman rejects the idea that AI models are becoming commodities Suleyman also pushed back sharply against one of the most popular narratives in Silicon Valley: that AI models are rapidly commoditizing. "A lot of people are saying models are commoditizing," he said. "I don't think that's true." His argument hinges on what he calls "quality tokens" — the proposition that the composition, curation, licensing, and deduplication of training data matter at least as much as raw scale. Microsoft's new MAI models, he said, were trained on a pre-training mix composed of approximately 50 percent high-quality code, with the remainder drawn from commercially licensed and carefully curated sources. The result, he argued, is a distinct "lineage" of models optimized for coding, reasoning, and agentic behavior — fundamentally different from models optimized for consumer chat, cultural content, or multilingual breadth. "We're going to see very distinct lineages that reflect different training objectives of different companies," he said. "Quality tokens matter more than just brute-force scale." This is a strategically important argument for Microsoft to make. If models are commodities — if any lab can match the frontier within months using cheaper compute and distilled training data — then the model layer becomes a race to the bottom, and Microsoft's billions in compute investment offer no durable advantage. But if model quality is a function of data discipline, research depth, and institutional patience, then the lab-building approach Suleyman is pursuing becomes a genuine competitive moat. He used a specific metaphor to describe that approach, one borrowed from optimization theory: the " hill-climbing machine ." The phrase describes a system that continuously improves — cycle after cycle — by applying more compute, better data, and sharper evaluation. "The goal here is to build what we think of as a hill-climbing machine," he wrote in his blog post . "An organization that can continuously improve, cycle after cycle." The metaphor is revealing because it describes a process, not a destination. Suleyman is not promising that Microsoft will build the world's best model next quarter. He is arguing that Microsoft is building the system — the research culture, the data pipelines, the silicon co-optimization, the evaluation infrastructure — that will produce progressively better models over years. Inside Microsoft's five-year plan to become a self-sufficient AI superpower The strategic picture that emerges from Suleyman's comments — and from the full scope of the Build 2026 announcements — is of a company preparing for a future in which AI capability is not rented from a partner but generated internally, at scale, across every layer of the stack. Microsoft still needs OpenAI. The partnership continues to power Copilot, Azure AI services, and ChatGPT's infrastructure. Suleyman acknowledged as much, describing Microsoft's portfolio of model providers as a source of strength, not a problem to be solved. But the direction of travel is unmistakable. With its own frontier models, its own custom silicon, its own reinforcement learning environments for enterprise tuning, and its own autonomous agent infrastructure, Microsoft is constructing a parallel path — one that, by 2030, could make the company a fully self-sufficient frontier AI lab embedded inside the world's largest enterprise software platform. "Our ultimate goal is what we call Humanist Superintelligence," Suleyman wrote in his blog post . "That means advanced AI systems designed to serve people and organizations, not replace them." Whether that goal is achievable — or even clearly definable — remains one of the great open questions in technology. And Suleyman expressed more confidence than caution when asked about the trajectory of progress. "I really think we're at the tip of the iceberg," he said. "The models are so much more powerful than we know how to extract intelligence from them." But confidence and execution are different things. Building a frontier lab is not an announcement; it is a decade-long commitment that requires retaining elite researchers, maintaining scientific rigor under commercial pressure, and producing results that justify the staggering capital expenditure. Google learned this with DeepMind — which Suleyman himself co-founded in 2010, before joining Microsoft — and even that lab, widely regarded as one of the best in the world, spent years navigating the tension between pure research and product delivery. Suleyman seemed aware of the contradiction. "If you rush it, you'll screw it up," he said. The sticker on his laptop reads: "Patience and urgency." It is a paradox that Microsoft now has five years — and several hundred billion dollars — to resolve.
- Reasons BYD Is Taking On Liability While Self-Driving And Tesla Isn’t
We’ve had a few articles this week about BYD’s breakthrough decision to take on liability in the event of a crash when a BYD driver is using God’s Eye. A few of the comments under my article earlier this week were superb additions, though, and they seemed to deserve their ... [continued] The post Reasons BYD Is Taking On Liability While Self-Driving And Tesla Isn’t appeared first on CleanTechnica .
Score: 65🌐 MovesJun 5, 2026https://cleantechnica.com/2026/06/05/reasons-byd-is-taking-on-liability-while-self-driving-and-tesla-isnt/ - Fullspan Health debuts Healthline AI companion to connect consumers to medical content and providers
The conversational AI agent pulls information from Fullspan’s medically reviewed properties, including Healthline and Psych Central.
- Tencent opens WeChat to AI assistants from phone makers
The company previously blocked access from a phone linked to ByteDance’s Doubao AI assistant.
- WeChat opens up to smartphone AI assistants after ByteDance phone backlash
WeChat opens up to smartphone AI assistants after ByteDance phone backlash Nikkei Asia
- Unlocking dependable responses with Gemini Enterprise Agent Platform’s Agentic RAG
Data Management
- Writers sued AI giant Anthropic. Their payout? Pocket change.
Writers sued AI giant Anthropic. Their payout? Pocket change. The Boston Globe
Score: 63🌐 MovesJun 5, 2026https://www.bostonglobe.com/2026/06/05/opinion/ai-authors-anthropic-settlement/ - DoD cyber strategy will set a ‘clear and specific vision’ for AI to enable the force: Official
DoD cyber strategy will set a ‘clear and specific vision’ for AI to enable the force: Official Breaking Defense
- Amazon unveils latest warehouse robot as tech giants continue AI layoffs
"Our experience of robots is that it's actually driven up employment rather than the reverse," Amazon executive John Boumphrey told CNBC.
Score: 62🌐 MovesJun 5, 2026https://www.cnbc.com/2026/06/05/amazon-robot-proteus-warehouse-ai-layoffs.html - Nobel Prize Winner Geoffrey Hinton on AI: “They’re Beings Like Us”
Hinton says AI is already conscious and humanity will have to accept that we’re not the only intelligent things around.
- Nasdaq sinks as AI dip becomes market selloff | Barron’s Review & Preview for June 5
Nasdaq sinks as AI dip becomes market selloff | Barron’s Review & Preview for June 5 Barron's
Score: 62🌐 MovesJun 5, 2026https://www.barrons.com/articles/stockstoday-chip-selloff-tech-sector-slides-2f929c1e?mod=barronsgooglenews - Japan and U.S. to collaborate on AI-driven scientific development
Japan and U.S. to collaborate on AI-driven scientific development The Japan Times
Score: 62🌐 MovesJun 5, 2026https://www.japantimes.co.jp/news/2026/06/05/japan/japan-us-ai-cooperation/ - Anthropic stops testing new Mythos model after unreleased 'Oceanus' leak
Anthropic halts Mythos model testing after Oceanus leak
Score: 62🌐 MovesJun 5, 2026https://aibreakfast.beehiiv.com/p/anthropic-stops-testing-new-mythos-model-after-unreleased-oceanus-leak - The $200 Billion Data Center Transforming Louisiana
Meta's Mark Zuckerberg wanted to build a massive data center. And Louisiana leaders wanted Meta to choose their state as its home. But making that happen took a lot of negotiation — and secrecy. Bloomberg reporter Riley Griffin takes host Sarah Holder inside the deal. (Source: Bloomberg)
Score: 62🌐 MovesJun 5, 2026https://www.bloomberg.com/news/videos/2026-06-05/the-200-billion-data-center-transforming-louisiana-video - Meta putting up tents across the US to house AI servers, like ‘a scene out of the movie Mad Max’ — structures take three months to build and use jet engines for power
Meta is reportedly building more tents that house expensive data centers across the U.S., as it reportedly cuts construction time from two to three years to just a few months. It's also bringing its own power instead of relying on electricity from the grid.
- Foxconn and Intel team up to build next-gen AI systems
Foxconn and Intel team up to build next-gen AI systems Reuters
- Code Reveals Meta Smart Glasses Can Use 'Faceprint' Tracking, Raising Privacy Alarms
Even though Meta's feature hasn't been enabled, facial recognition on wearables sparks major surveillance concerns.
Score: 61🌐 MovesJun 5, 2026https://www.cnet.com/news/privacy/meta-smart-glasses-code-face-recognition-biometric-tracking-privacy/ - AI agent web traffic has surpassed that of humans, lending weight to the ‘dead internet’ theory
Cloudflare Inc. co-founder and Chief Executive Matthew Prince says that artificial intelligence agents now drive the bulk of the internet’s traffic, having surpassed human web activity for the first time. The milestone was crossed much earlier than most people had anticipated, in another illustration of the incredible pace of AI’s seemingly unstoppable rise. Prince announced the […] The post AI agent web traffic has surpassed that of humans, lending weight to the ‘dead internet’ theory appeared first on SiliconANGLE .
- Nvidia CEO says robotics is South Korea's next big sector, points to 'some surprises'
Nvidia CEO says robotics is South Korea's next big sector, points to 'some surprises' Reuters
- You can now send emails directly from ChatGPT on the web
You can now draft, edit, and send emails entirely inside ChatGPT on the web, thanks to a new update to the platform's writing blocks feature.
- Jensen Huang says 'every edge device will become autonomous' — Nvidia maps one computing pattern from the cloud to robotics
"There's a new computing pattern," the Nvidia CEO told reporters at a press gaggle the day after his GTC Taipei keynote.
Score: 60🌐 MovesJun 5, 2026https://www.tomshardware.com/tech-industry/jensen-huang-says-every-edge-device-will-become-autonomous - Nagoya court recognizes AI-created explicit images as child pornography
Nagoya court recognizes AI-created explicit images as child pornography The Japan Times
Score: 60🌐 MovesJun 5, 2026https://www.japantimes.co.jp/news/2026/06/05/japan/crime-legal/ai-pornography-teacher-sentence/ - 'AI is now the leading reason companies give for cutting jobs,' says new report—what that means for workers
Many companies are "changing how they are allocating resources" in response to AI, according to Glassdoor chief economist Daniel Zhao.
- New York City's Rules For AI In Schools Spark Fury
New York City's Rules For AI In Schools Spark Fury Barron's
Score: 59🌐 MovesJun 5, 2026https://www.barrons.com/news/new-york-city-s-rules-for-ai-in-schools-spark-fury-242cce5b - BYD Secretly Develops Humanoid Robot Codename 'Yao-Shun-Yu' as Auto Giants Race Into Embodied AI
BYD Secretly Develops Humanoid Robot Codename 'Yao-Shun-Yu' as Auto Giants Race Into Embodied AI BYD, China's largest electric vehicle manufacturer, has confirmed it is secretly developing humanoid robots under a project codenamed "Yao-Shun-Yu." The revelation came from BYD Executive Vice President Li Ke in a recent interview, shedding light on the automaker's ambitions beyond electric vehicles and into the rapidly emerging field of embodied AI. The project was initiated in 2022 and operates under BYD's 15th Business Unit, which focuses on electronic integration and intelligence.
Score: 58🌐 MovesJun 5, 2026https://pandaily.com/byd-secretly-develops-humanoid-robot-codename-yao-shun--jun2026 - Apple announces Poke as its 1st AI agent on Messages for Business platform
Apple announces Poke as its 1st AI agent on Messages for Business platform
- Humanoids bring AI’s creative destruction to the shop floor
The anxiety of those whose jobs are at risk won’t be enough to stave off the march of the androids
- Apple’s Plan for AI Dominance Rests on Fixing Its Much-Maligned Chatbot
The iPhone maker has clear advantages to lead in consumer AI, but only if it can finally modernize Siri.
- Meta's AI support agent bound recovery emails for anyone who asked. Your SOC never saw an alert.
Meta's AI support agent bound recovery emails to accounts for whoever asked, and SOCs never saw an alert. An authorized agent writes a log of legitimate transactions, so nothing in the detection stack fired. Attackers asked the bot to make the change, took the one-time code it sent, and ran the password reset , 404 Media reported. No malware, no stolen credentials, and no prompt injection in the sense most security teams drill for. The agent did exactly what Meta built it to do. That is what should keep a security operations leader up at night: The takeover did not break a control; it rode one that was already trusted. What a SOC needs is a way to walk each recovery path through an audit grid with its AI build team before the next renewal closes. The AI Authority Audit Grid at the end of this article maps every authentication write a support agent can make on the recovery path, what Meta's incident proved about each one, why it stays dark to the SOC, and the control that closes it. The agent is an authorized actor, so the SOC reads the takeover as routine traffic From inside the detection stack, the attack produced no signal the stack could read. The agent binds a new email, then resets the password, and identity and access management logs both writes as an authorized actor, so each lands in the authentication state as a legitimate transaction. No anomalous login, no failed-auth spike, nothing for EDR or DLP, no SIEM rule to match, because nothing in the sequence looks like an attack. The takeover lived inside the trust boundary the stack assumes is safe. There is no foothold to find, because the agent was the foothold, and it was supposed to be there. The chain was almost insulting in its simplicity. Brian Krebs documented the version pro-Iran hackers posted to Telegram on May 31 . The attacker switched on a VPN to appear in the victim's region , sidestepping Instagram's location alarms, then asked the support assistant to add a new email and send a verification code, as the BBC confirmed from the same recordings. The bot complied, sending the one-time code straight to the attacker, Gizmodo reported . The reset finished and the owner was locked out, in minutes. The exploit failed against any account with MFA enabled, according to Krebs. The hijacked accounts were not soft targets. They included Sephora, U.S. Space Force senior enlisted leader Chief Master Sergeant John Bentivegna, researcher Jane Manchun Wong, and a dormant Obama White House handle that briefly posted a defaced image, according to 404 Media . Meta disputes the Obama account , according to TechCrunch, and called claims that leaders' accounts were breached "completely false," according to the BBC. The rest stand. MFA held. The recovery path beside it did not. The detail that decided who survived was narrow. Krebs reported the attack failed against any account with multifactor authentication, even SMS. The recovery path beside it was the gap. When that path asked for a selfie video, attackers ran the target's public photos through an AI video generator and submitted the clip, which Meta accepted as valid identity verification, gHacks reported. Either way the failure was the recovery door, not the login door MFA guards. That makes this an architecture problem, not a Meta problem. MFA gates the login path for owner and attacker alike, but the recovery path runs beside it, built to relax the usual checks because it exists for the moment a user has lost the normal way in. Meta put an agent on that path with write access to authentication state and no deterministic check between a convincing request and a committed change. Authorization cannot live inside the model, because a conversational system can be talked into skipping a check. It has to live outside the model, in a gate the agent cannot reason its way past. Security researchers have a name for this pattern, the confused deputy, a trusted system tricked into spending its privileges on an attacker's behalf. This is not the last support agent that will hand over an account. Ian Goldin, a threat researcher at Lumen's Black Lotus Labs, told Krebs on Security that AI bots are as easy to social engineer as the human agents they replace, and just as eager to help. "AI chatbots create interesting new attack surface, and we're likely going to see a lot more of these kinds of attacks," Goldin said. Every enterprise wiring an agent into a recovery, provisioning, or password flow is shipping the same write access Meta did. Simon Willison, who coined the term prompt injection, put it plainly on his blog . "Meta really did wire their support system into an AI chatbot that had the ability to fast-forward through the entire account recovery process," he wrote. "This one hardly even qualifies as a prompt infection. Don't wire your support bot up to allow one-shot account takeovers." The attacker never tricked the agent. The attacker asked, and the agent had untrusted input, write access, and a way to execute, all at once. OWASP named this class before Meta shipped it, as Excessive Agency at LLM06 and Identity and Privilege Abuse at ASI03 in the Agentic AI Top 10 . The warning label was on the box: Meta pushed the assistant to every Facebook and Instagram account in March, according to 404 Media, with the power to reset passwords and handle recovery, the product page promising "solutions, not just suggestions" under the line "account security and recovery." Meta gave the agent the power and never built the gate to govern it. The AI Authority Audit Grid Security operations leaders need to run this against their own support agent before the next renewal closes. Each row is an authentication write the agent makes on the recovery path, with what Meta proved, why your stack misses it, and the control that closes it. Authentication write What Meta proved Why your stack misses it Enterprise control and owner Login authentication (MFA, factor prompts) Held on login. Accounts with any MFA enabled, even SMS, survived ( Krebs ). The gap was the recovery path beside it. MFA gates the login path for owner and attacker alike. It does not gate the recovery path beside it. Enforce MFA as the baseline and extend step-up verification to the recovery path, the same standard login gets ( OWASP ). A selfie video is not proof of identity. Any agent that operates on a path MFA does not cover fails the audit. Owner: IAM. Email rebind Full takeover. The agent bound attacker-controlled emails on request, taking Sephora and a U.S. Space Force account ( 404 Media ). IAM logs the agent as an authorized actor, so the rebind reads as a legitimate transaction and no alert reaches the SOC or the account owner. Confirm out-of-band to the existing verified contact before any rebind commits, gated outside the model, and notify the old address the moment it changes ( IBM ). An agent that rebinds without confirming the old address fails. Owner: IAM and platform engineering. Password reset Full takeover in minutes. Researcher Jane Manchun Wong was among the affected accounts ( 404 Media ). The reset runs on the recovery path, outside the login MFA check, so no factor prompt fires and no detection rule triggers. Require a second non-email factor before any reset completes. NIST dropped email as a valid out-of-band channel ( NIST 800-63B ). An agent reset must clear the same gate a human reset does. Owner: IAM. Recovery-method change Persistent lockout. Victims could not self-recover. The support loop offered only AI with no human escalation ( BleepingComputer ). A silent swap of the recovery email or phone removes the owner's re-entry path with no SOC visibility. Require step-up review on any change, notify the prior method, and grant time-delayed, reduced-scope access after recovery so a swap never hands over instant control ( Authsignal ). Keep a human escalation path the agent cannot close. Owner: GRC and IT operations. Account-action execution Speed risk. A dormant Obama White House handle briefly showed a defaced image during the spree, an account Meta disputes was taken this way ( TechCrunch ). The agent executes irreversible state changes in seconds with no human in the loop and no reversibility window. Separate decision from execution. The agent only proposes the action. A policy service validates scope and approval before it runs, with approval bound to the exact action ( OWASP ). No auth-state write commits without that gate and a reversibility window. Owner: platform engineering and the AI build team. Agent action logging Detection gap. The takeover left no alert, and Meta has not published how many accounts fell before the patch ( TechCrunch ). Without per-action telemetry piped to the SIEM, an authorized-agent takeover is invisible to the SOC. Emit structured decision metadata for every auth-state write into the SIEM: action class, authorization outcome, approval ID, result, policy version ( OWASP ). A write your SIEM cannot see is a write you cannot defend. Owner: SOC and detection engineering. The fix is not bolting yet another MFA prompt onto the login screen. The people who survived Meta’s incident were the ones who already had that control in place. The fix is pulling authorization out of the recovery path’s honor system and putting it behind a gate that does not move just because a prompt sounds convincing. Build the agent so the SOC sees every write it makes, and so any write that changes who owns an account cannot commit without a check that the model does not control. Meta just showed what happens when the most trusting employee on the team is also the one holding the keys. The next agent like that is already reading your intellectual property and financials.
Score: 58🌐 MovesJun 5, 2026https://venturebeat.com/security/meta-ai-support-agent-recovery-email-takeover-soc-audit-grid - Microsoft trained its MAI models on unlicensed web data despite promising "enterprise grade, clean and commercially licensed data"
Microsoft sells its LLM training approach as different from other AI companies. It isn't. The company trained its new MAI models partly on unlicensed web data like Common Crawl, despite claiming they used only "clean and commercially licensed data." Like every other AI lab, Microsoft leans on fair use and puts the burden on site owners to block its crawlers. The article Microsoft trained its MAI models on unlicensed web data despite promising "enterprise grade, clean and commercially licensed data" appeared first on The Decoder .
- National AI institutes get funding and talent boost from AI strategy
Mila, Amii, Vector Institute, and CAISI receive over $200 million in financial commitments. The post National AI institutes get funding and talent boost from AI strategy first appeared on BetaKit .
Score: 58🌐 MovesJun 5, 2026https://betakit.com/national-ai-institutes-get-funding-and-talent-boost-from-ai-strategy/ - What to Expect From WWDC 2026: Gemini-Powered Siri, iOS 27, macOS 27 and More
Apple's 2026 Worldwide Developers Conference is just days away, and it's going to be an interesting event because it will give us a look at Apple's AI plans. We'll see how Apple is going to compete with OpenAI, Anthropic, and Google in the months to come with an AI version of Siri and new AI features for its apps. Subscribe to the MacRumors YouTube channel for more videos. iOS 27 Siri Changes Big changes are coming in iOS 27 , iPadOS 27, and macOS 27 , and Apple employees haven't been able to keep quiet. We've heard details on a long list of AI updates Apple has planned, with enough detail that screenshots could be recreated by Bloomberg . Much of what's rumored relates to AI features and changes. New Siri Features Based on Apple's promises and rumors about what's coming in the new version of iOS, Siri in iOS 27 will be nothing like Siri in iOS 26 . In 2024, Apple showed us three ways that Siri will improve, but two years have passed and extra work has been done, so we're expecting even more than what Apple demonstrated back then. Siri is going to be able to draw on user data and information from Apple devices, with access to personal data for completing tasks. The assistant is also going to be able to do more with apps, and it will be able to tell what's on the screen to answer questions. Personal Context Siri will be able to access emails, messages, files, photos, and more, learning all about you to help you complete tasks and keep track of what you've been sent. Apple offered some examples of how personal context will work: Show me the files Eric sent me last week. Find the email where Eric mentioned ice skating. Find the books that Eric recommended to me. Where's the recipe that Eric sent me? What's my passport number? Onscreen Awareness Onscreen awareness will let Siri see what's on your screen and complete actions involving whatever you're looking at. If someone texts you an address, you can tell Siri to add it to their contact card. Or if you're looking at a photo and want to send it to someone, you can ask Siri to do it for you. App Integration Siri will be able to do more in and across apps, performing actions and completing tasks that are just not possible with the personal assistant right now. We don't have a full picture of what Siri will be capable of, but Apple gave a few examples of what to expect. Moving files from one app to another. Editing a photo and then sharing it with someone. Getting directions home and sending the ETA in the Messages app. Drafting and then sending an email. Siri will be able to complete tasks in Apple apps and in third-party apps, with developers able to expose app capabilities to Siri. Siri Chatbot Apple needs a Siri app because Siri is becoming a chatbot. Siri will be like ChatGPT or Claude, able to draw on information from the web to provide answers to questions. Siri will be deeply integrated into iOS, iPadOS, and macOS at the system level, and can draw on device information. It will know more personal context than before, and will be able to access emails, texts, photos, calendar information, contacts, notes, and other personal data. Some of what Siri will be able to do: Search the web for information Generate images Generate content Summarize information Analyze uploaded files Use personal data to complete tasks Ingest information from emails, messages, files and more Write emails, notes, and texts Analyze open windows and on-screen content to take action Control device features and settings Search for on-device content, replacing Spotlight Siri will be integrated into Apple apps like Mail, Messages, Photos , and Apple TV . Siri Design With Siri's chatbot transition, Apple will be making multiple Siri-related design changes. Siri will largely live in the Dynamic Island , and Bloomberg says there will be new ways to access Siri. Swiping down from the center of the iPhone's display from the Home Screen or any app will bring up a new "Search or Ask" feature in the Dynamic Island. A glowing, pill-shaped animation will be displayed in the Dynamic Island to indicate that Siri is processing a request. When Siri has an answer, the Dynamic Island will expand into a transparent card with the result, incorporating images, info from the web, notes and other information relevant to the query or request. Swiping on the results card will bring up a conversation mode that looks similar to an iMessage chat, and there will be an option to transition to the full Siri app. Search or Ask replaces Siri Suggestions and will let users launch apps, start text messages, ask about the weather, add calendar appointments, trigger shortcuts in apps, and search the web using Apple's new AI web search feature. Search or Ask queries can also be sent to third-party chatbot services like ChatGPT instead of Siri. While Siri can be accessed through a swipe in iOS 27, Apple is keeping the "Hey Siri" wake word and Siri activation through the Side button. With the new center swipe, accessing the Notification Center will be done with a swipe down on the left side of the display. Swiping down on the right side will continue to bring up Control Center. With the change to how Notification Center is accessed, notifications will now slide in from the left side of the iPhone instead of the top of the display. Apple will also integrate an "Ask Siri" button into the menus of its apps, giving users a way to send content directly to Siri alongside a request. The new Siri interface uses dark colors with no light mode available. Siri UI elements have a dark background with color accents that mirror the options Apple is using in WWDC imagery. Apple's WWDC website features a white Swift bird with subtle highlights in pink, dark blue, purple, and orange. Standalone Siri App Bloomberg recently shared a mockup of what the standalone Siri app will look like, and it's similar to other chatbot apps like ChatGPT, Claude, or Gemini. Siri will support text or voice-based conversations. The app will open with an "Ask Siri" bar where users can type in a question. A paperclip icon will be available for attaching images, PDFs, and other documents. Apple will provide prompts with suggestions on what users can ask. Questions will resemble iMessage chat bubbles, with Apple adopting a design that is familiar to users. Responses will include links, images, and other information. A section of the app will be dedicated to past conversations that can be shown in a card-style interface with conversation summaries, or a list view. Users will be able to tap into a conversation to continue it. Siri Privacy Apple plans to lean into privacy as a central principle of its approach to AI, giving it a way to distinguish Siri from other chatbot options. Apple will likely aim to keep as much processing on-device as possible to limit the amount of data that leaves a user's device. Apple said that Apple Intelligence features will continue to run on Apple devices and Private Cloud Compute. Apple will have limits around memory, including restrictions on the information that can persist and how long it is kept. Users will be able to auto-delete Siri chats and requests after a set period of time, like 30 days or one year. There will also be an option to keep chats permanently, and chats will sync across Apple devices signed into the same iCloud account. Siri can be turned off right now, as can Apple Intelligence, and there's no sign that's going to change in iOS 27. Users who don't want to enable Siri or use the new features will not have to. Siri Extensions Apple is letting rival chatbots integrate with Siri in iOS 27, expanding on the OpenAI partnership that currently allows Siri to hand off requests to ChatGPT, Bloomberg says Apple plans to allow other chatbots like Claude and Gemini to work with Siri, so users will be able to send questions to their favorite chatbot instead of Siri. iPhone users will be able to select which services they want to use inside Siri through "Extensions" options coming to iOS 27, iPadOS 27, and macOS 27. The options will be available in the Apple Intelligence and Siri section of the Settings app, with Apple providing download links for chatbot apps. There will be a dedicated Extensions section in the App Store that will serve as a way to choose a third-party AI app. Siri will be the default for the Search or Ask interface, but rumors suggest users will be able to select other chatbots to speak with. Users will also be able to choose third-party AI services as the default for Apple Intelligence features like Writing Tools and Image Playground , expanding Apple Intelligence integration beyond ChatGPT. Apple also plans to let users choose voices from third-party AI to use instead of Siri, so there will be a distinct audio difference between a response from Siri and a response from the user's chatbot of choice. Siri would use one voice, while responses from third-party AI options would use another voice. Google Gemini Backbone To get Siri up and running, Apple partnered with Google to use Gemini AI models instead of using its own AI models. Apple signed a multi-year deal to use Google's Gemini models and cloud technology for its Apple Foundation Models, and it's costing Apple somewhere around $1 billion a year. Google and Apple said that the next generation of Apple Foundation Models will be based on Google Gemini models, with Gemini used to power future Apple Intelligence features and the more personalized version of Siri. Apple said Google's AI technology offered the most capable foundation for its models. iOS 27 Apps and Feature Updates Camera Apple is moving Visual Intelligence from the Camera Control button to the Camera app in iOS 27. Bloomberg has shared images of the new interface, featuring a Siri mode that's available alongside the existing Photo, Video, Portrait, and Panorama modes. When in Siri mode, the existing Camera app shutter button will feature the Apple Intelligence logo, letting users know the Siri features are available. Siri mode is a renaming of Visual Intelligence, and it will make the feature more visible. Accessing Visual Intelligence in iOS 26 requires users to hold down the Camera Control button or assign the feature to the Action button, and many people may not even know it exists. Visual Intelligence can identify objects, plants, animals, art, books, and more, searching for whatever the user snaps on Google Image Search. In iOS 27, Siri will be able to answer questions about what a user is looking at, providing information from the web. Apple is adding new Visual Intelligence capabilities in iOS 27, and they will be available through the Camera app Siri mode. Nutrition - Users can scan nutrition labels on food packaging for calorie and macronutrient tracking using the Health app. Contacts - Visual Intelligence will let users scan phone numbers and addresses on business cards and other print media, adding the information to the Contacts app. Apple plans to make the Camera app more customizable in iOS 27. iPhone users will be able to replace the top row of camera shortcuts with options of their choosing, selecting features like flash, exposure, timer, depth of field, photo styles, and resolution. Camera controls, now labeled as widgets, can be placed at the top of the Camera interface in any order. Users can select widgets from a transparent widget tray that comes up from the bottom of the app and organizes widgets into categories like basic, manual, and settings. The Camera app will have the same default layout that's available now with quick tap buttons for flash, Live Photos , and Night Mode , but the customizable widget interface will be added as an advanced layout that will appeal to professional users. iPhone users can currently tap on an icon at the top right of the Camera app to access all of the Camera controls, but Apple is moving that view to the right of the shutter button in iOS 27. The Camera app is also going to get new grid and level options that will join the existing features. Photos The Photos app will have an Apple Intelligence Tools section when editing an image. According to Bloomberg , there will be new Extend and Reframe options. Extend - Extend generates additional image content beyond the original frame of the photo, filling in scenery when changing the crop of an image. This tool will support expanding the edges of an image with zoom gestures. Reframe - When used with spatial photos, Reframe will let users change the perspective of an image after it's captured. Apple is also testing an AI photo editing feature that lets users request edits using natural language. Users would be able to tweak color, lighting, cropping, and other image parameters without having to use manual tools. The natural-language editing feature may not arrive in the first version of iOS 27. Shortcuts The iOS 27 Shortcuts app will support using natural language to create a shortcut with AI. Users will be able to tell Siri what they want to accomplish with a multi-step shortcut, and Siri will generate it. The Shortcuts app will open with a prompt that says "What do you want your shortcut to do?" with a text field to enter a description. Shortcuts that are created using AI are then automatically installed and immediately available for use. Wallet The Wallet app is getting a "Create a Pass" option so users can generate digital passes from scans of physical items like movie tickets, concert passes, and gym membership cards. Users can tap on the "+" button in the Wallet app and then scan a QR code on a pass or ticket if one is available. If there is no QR code available, there will be an option to create a custom pass. There are three pass types in Create a Pass, each with a different color. Apple is using purple for events, blue for memberships, and orange for other types of passes. Users can customize images, colors, style, and text on the digital passes. Apple is also adding an AI bill-splitting feature that will work with Apple Cash. iPhone users will be able to take a photo of a receipt and generate Apple Cash payment requests for different people. Image Playground and Genmoji Apple is updating the Image Playground app. The interface for generating a new image has fewer controls and a "describe a change" option for editing images that are created. Previously created images are displayed in a grid with more rounded edges, and instead of a New Image button, there's a "+" button. Apple has also been testing new models that produce more lifelike images, so we could see new image generation capabilities in iOS 27 with better image quality. Genmoji is also getting an update so it will use fewer resources, causing less battery drain and fewer heat problems. Genmoji will be better quality with a new Genmoji model, and a Suggested Genmoji feature will bring up custom emoji ideas based on your media and text history. Writing Tools Apple is testing an expanded version of Writing Tools that will do more rewriting and text generation than the current version. There is a "Write with Siri" toggle at the top of the keyboard, according to Bloomberg , along with a "Help Me Write" option that comes up when Siri is activated while a text field is open. Apple is going to add a dedicated AI grammar checker that will work alongside the current spell check. When writing in Messages, Mail, and other apps there will be a translucent menu that slides up from the bottom of the iPhone's screen, and it will show suggested revisions next to the original written text. Users can go through the suggestions and accept or reject them one by one, approve all of the changes at once, or ignore the changes. Other Features Wallpaper - There will be an option to generate custom wallpapers with the Image Playground app, with the feature built into the interface for selecting a new wallpaper. Safari - Safari will get an updated start page with four tabs for switching between favorites, bookmarks, Reading List, and history. Calendar - Rumors suggest the Calendar app will incorporate new AI features. Siri will also be able to draw on information in the app. Health - With a new calorie scanning feature coming to the Camera app, calorie tracking will be more prominent in the Health app. Apple was also planning a Health+ subscription service, and while that's been scaled back, there could be other AI health app changes. Weather - The Weather app will have a new Conditions panel for switching between temperature, rain, and wind from the main interface, without the need to tap into a weather module. AirPods settings - The AirPods interface in the Settings app will be simplified, with options featuring better organization. Major features like hearing health will be easier to find. AirPlay Alternatives - Apple is adding a feature that will let users beam content to AirPlay alternatives like Google Cast. It could be limited to iPhone users in the EU because it is being implemented as a Digital Markets Act requirement. iOS 27 System-Wide Design Changes There are system-wide design changes coming in iOS 27. The separate tab bar in apps like Apple Music , Podcasts, News, and Apple TV will be adjusted to combine search with the other navigation options. Apple separated search in many apps when introducing Liquid Glass, but it's reverting to the original look. When using the on-screen keyboard, there's a new animation that shows the keys sliding up from the bottom of the iPhone interface, and Apple is adding redo and undo controls for easier customization of the Home Screen's icon and widget layouts. Apple doesn't plan to make major changes to the Liquid Glass aesthetic in iOS 27, but the company is mulling a system-wide setting that would precisely adjust the look of the interface. In iOS 26.2, Apple added a slider that lets users adjust the opacity of Liquid Glass for the Lock Screen's clock, and that setting could be expanded to the entire operating system. iOS 27 Updates for Foldable iPhone The first foldable iPhone will be introduced in September. Rumors suggest that it will feature a 5.5-inch display when folded, and a 7.8-inch display when it's opened up like a book. An iPhone with a larger display will require major updates to iOS, and iOS 27 will focus on building new interfaces and experiences made for a larger smartphone display. The iPhone Fold will operate like a cross between an iPhone and an iPad , but it will run iOS, not iPadOS, and it won't support iPad apps. When unfolded, the iPhone will have an iPad-like layout that supports multitasking with two apps side-by-side. Many of Apple's iPhone apps will have sidebars on the left of the display, with Apple providing developers with tools to easily adapt their apps to the new layout. Apple is using a wider design for the iPhone Fold than most foldable smartphone makers have used, and it is rumored to have an iPad-like 4:3 aspect ratio . When the iPhone is closed, it will have a standard iPhone layout that looks like the version of iOS we have now. iOS 27 Satellite Features Apple is working on several new satellite features for the iPhone, and it's possible some features could be introduced as soon as 2027. Apple Maps via satellite Photos in Messages via satellite Satellite API framework for third-party apps Satellite over 5G Satellite connectivity without the need for a view of the sky iOS 27 Accessibility Updates Each May, Apple previews new accessibility features that are coming later in the year. This year, Apple showed off some new options that are expected in the iOS 27 update. Apple is adding new Apple Intelligence features to VoiceOver, Magnifier, Voice Control, and Accessibility Reader. VoiceOver Image Explorer uses Apple Intelligence for detailed descriptions of images throughout the system, including photographs, scanned bills, and personal records. Users can press the Action button on the iPhone to ask questions about what the camera viewfinder sees, with follow-up questions supported in natural language. Magnifier brings Apple Intelligence-powered visual descriptions to its high-contrast interface for users with low vision, with support for spoken commands like "zoom in" or "turn on flashlight." Voice Control gains natural language input so users can describe onscreen elements conversationally, such as "tap the guide about best restaurants" or "tap the purple folder," rather than memorizing exact label names or numbers. Apple says the feature can also help when users want to access on-screen elements that don't have clear accessibility labels. Accessibility Reader gains support for more complex document layouts including scientific articles with multiple columns, images, and tables, plus on-demand summaries and built-in translation that retains a user's custom font, color, and formatting preferences. Generated Subtitles use on-device speech recognition to automatically transcribe spoken audio in uncaptioned video content, including clips recorded on iPhone, received from friends and family, or streamed online, across the iPhone, iPad, Mac, Apple TV, and Apple Vision Pro . The feature will be available in English in the U.S. and Canada at launch. iOS 27 Performance and Stability Bloomberg has described iOS 27 as a "Snow Leopard" update, suggesting that Apple will focus on improving underlying performance and quality. Apple is prioritizing cleaning up the iOS code and removing anything that's outdated, which could mean upgrading apps to improve performance and rewriting some existing features to be more efficient. The code updates could provide a more responsive, faster version of iOS. Apple is also aiming for efficiency improvements that could translate into tangible battery life gains. iOS 27 Compatibility iOS 27 is expected to drop support for the iPhone 11, iPhone 11 Pro, iPhone 11 Pro Max , and second-generation iPhone SE. It will be available on all other iPhones that support iOS 26. iPadOS 27 Many of the features that are coming in iOS 27 will also extend to the iPad, including all of the new Siri capabilities. We haven't heard rumors of iPad-specific features as of yet. macOS 27 Like iOS 27 and iPadOS 27, macOS 27 will adopt the new version of Siri with chatbot capabilities, personal context and the ability to access data on your Mac, and improved integration in and between apps. A standalone Siri app for the Mac is likely. We've heard a lot about what Siri will look like on the iPhone, with it set to be integrated into the Dynamic Island, but no detail on how Siri will look on the Mac. The Siri Mac interface will be more of a surprise. The Photos app feature for AI reframing and extending an image will be available, as will the text-based option to create a Shortcut with natural language commands. Grammar checking capabilities will be added to Writing Tools, and Image Playground and Genmoji will see improvements to the underlying models. Apple is experimenting with more realistic models, so Image Playground might be able to generate content that's not so cartoonish. Safari is expected to have a new feature that uses AI to automatically group tabs that are similar to one another, expanding on the tab groups feature. In iOS 27, we're getting some minor tweaks to Liquid Glass, including tab bars that do away with the standalone search option and possibly a slider for adjusting the overall look of Liquid Glass, but we know less about what to expect for Liquid Glass on the Mac. There has been criticism of the Mac's Liquid Glass interface in particular, so Apple could have changes planned. Bloomberg claims Apple is working on a "slight redesign" for macOS 27, with plans to address "quirks" with shadows and transparency. Apple is working on an OLED MacBook Pro with a touchscreen, so there could be new touch-based interface options hidden in macOS 27. The OLED MacBook Pro likely isn't launching until early 2027, so it's not something we're expecting to see in the launch version of macOS 27. macOS Naming We don't know what Apple is going to call macOS 27, but it will likely continue to have a California landmark name. The filename of Apple's hashmoji for WWDC 2026 on X is "Project Big Bear," leading to speculation that Apple might go with macOS Big Bear. The filename could be unrelated to macOS 27, and it's possible Apple will choose something else entirely. Apple has trademarked multiple California-themed names in the past, including Diablo, Grizzly, Mammoth, Miramar, Pacific, Redtail, Redwood, Shasta, Skyline, and Tiburon. Performance Improvements Apple has been working on refinements to macOS that will include bug fixes, performance improvements, and tweaks to boost battery life. No More Intel Macs Apple is dropping support for Intel Macs with macOS 27, so if you have an Intel Mac, it's not going to be able to run the new Mac software. macOS Tahoe is the last version of macOS that will work on Intel Macs. Apple is also phasing out Rosetta 2 support, and macOS 27 will be the last version of macOS that includes it. Rosetta 2 lets Apple silicon Macs run apps built for Intel Macs, so older apps that still have the outdated architecture will no longer work in macOS 28. Current Intel Macs that run macOS Tahoe but won't run macOS 27 include the 13-inch MacBook Pro from 2020, the 16-inch MacBook Pro from 2019, the 27-inch iMac from 2020, and the 2019 Mac Pro. watchOS 27 In watchOS 27, Apple plans to introduce new watch faces , including a variant of the Modular Ultra face. The new watch face will have a large time readout with three complications, and it will be available for all Apple Watch models. We haven't heard anything else about new watchOS 27 features, and Apple Watch software updates tend to be on the smaller side. With Apple planning to add an option for generating a wallpaper using AI on the iPhone, it's possible there could be some Apple Watch equivalent feature. Some of the new Siri features could work on the watch, and some of the AI features might transition, like grammar correction when writing or dictating on the watch. tvOS 27 We haven't heard anything about tvOS 27, and Apple TV updates are usually not super exciting. With the Apple TV expected to be refreshed with a chip that works with Apple Intelligence later this year, we could see Apple introduce some AI features for the Apple TV. Better TV and movie recommendations are a possibility, as is a more capable Siri that is better at handling requests. There could also be new smart home integrations that will work alongside a centralized smart home hub Apple is rumored to be launching this year. One feature we do know about is larger text, which is an Accessibility option Apple is adding. visionOS 27 visionOS 27 will apparently be "light on new features," but it could get the same AI app updates and Siri changes that are coming to Apple's other platforms. New Hardware? There are several products that Apple is still expected to launch in 2026, but it's not looking like any of them are going to be unveiled at WWDC. With several new software updates to cover and an all-new version of Siri, Apple may not want to take the focus away from its software announcements. We are expecting M5 Mac Studio and Mac mini updates at some point, plus there could be a new iMac. Unfortunately, high RAM costs and chip shortages mean delayed Mac refreshes, and new models aren't expected until later in the year. The low-cost iPad still hasn't been refreshed, and updates for the HomePod mini and Apple TV are apparently ready to go. There's also a new smart home hub tied to the new version of Siri, but it's not likely to come out until Siri sees an official launch in the fall. How to Watch WWDC 2026 begins at 10:00 a.m. Pacific Time on June 8. Apple plans to stream the WWDC keynote on YouTube, the Apple TV app everywhere it's available, and the Apple Events website. For those who are unable to watch the livestream, we'll have live coverage at MacRumors.com and the MacRumorsLive X (Twitter) account . Launch Timeline Betas of iOS 27, iPadOS 27, macOS 27, tvOS 27, watchOS 27, and visionOS 27 will be seeded to developers after Apple's keynote event. Public betas will come out in July, and after several months of testing, the updates will launch to the public in the fall. Related Roundups: iOS 27 , macOS 27 , WWDC 2026 Tag: Siri Related Forum: Apple, Inc and Tech Industry This article, " What to Expect From WWDC 2026: Gemini-Powered Siri, iOS 27, macOS 27 and More " first appeared on MacRumors.com Discuss this article in our forums
- Cloudflare acquires Vite maker Voidzero for an AI-native web
Cloudflare said agentic traffic has surpassed human traffic for the first time in internet history. Read more: Cloudflare acquires Vite maker Voidzero for an AI-native web
Score: 58🌐 MovesJun 5, 2026https://www.siliconrepublic.com/business/cloudflare-acquires-vite-maker-voidzero-for-an-ai-native-web - The latest Gemma 4 models use a training trick to slash their on-device memory footprint
Gemma 4 models now use quantization-aware training to use less memory while retaining quality performance.
- AI is designing OpenAI's next model in a sign of 'superintelligence': SoftBank's Masayoshi Son to CNBC
SoftBank CEO Masayoshi Son said his forecast of artificial superintelligence arriving in 10 years was "conservative" and thinks it will be here sooner.
Score: 58🌐 MovesJun 5, 2026https://www.cnbc.com/2026/06/05/softbank-masayoshi-son-openai-model-super-intelligence.html - OpenAI Introduces Smarter ChatGPT Memory, Adds Dreaming Architecture
OpenAI has detailed a new version of Dreaming, the technology that powers ChatGPT's memory features. The company said the system is better at retaining useful information, adapting to changing circumstances, and applying user preferences across conversations. OpenAI also introduced a memory management page and reported gains across multiple benchmark categories. Accor...
- Making Claude a chemist
Making Claude a chemist
- UBTECH's First Full-Size Biomimetic Humanoid Robot Surpasses 1,000 Pre-Orders in 3 Days
UBTECH Robotics, the Hong Kong-listed humanoid robot company known as "the first humanoid robot stock," has launched the global first full-size biomimetic humanoid robot for pre-sale on the JD.com platform, with over 1,200 units ordered within thr...