AI News Archive: June 1, 2026 — Part 4
Sourced from 500+ daily AI sources, scored by relevance.
- 8-day battery, AI health coaching and no screen: Fitbit Air review
The Fitbit Air offers health tracking, AI-powered coaching, and up to eight days of battery life without the distractions of a screen.
- Nobel Prize-Winner Demis Hassabis Says AI Job Cuts Are Dumb. Research Agrees
The DeepMind CEO says those cutting jobs because of AI ‘lack imagination.’ Here’s all the science suggesting he’s right.
- US AI unicorn Positron opens Dubai base at DIFC as demand for AI infrastructure surges
US AI unicorn Positron opens Dubai base at DIFC as demand for AI infrastructure surges Arabian Business
- With DGX Station for Windows, Nvidia squeezes 1 trillion-parameter AI supercomputer into a deskside form factor
Nvidia Corp. says it’s uprooting supercomputers from the vast, sprawling data center complexes they normally live inside and squeezing them into compact, desktop-sized workstations that can sit on or aside the desks of individual developers, researchers and data scientists. That’s the idea behind the new Nvidia DGX Station for Windows, which is said to be […] The post With DGX Station for Windows, Nvidia squeezes 1 trillion-parameter AI supercomputer into a deskside form factor appeared first on SiliconANGLE .
- Read the Frontier AI Trends Report
Report on frontier AI trends.
- HP OmniBook Ultra 16 (2026), OmniBook X 14 (2026) Unveiled With Nvidia's RTX Spark 'Superchip'
HP has announced that its next-generation HP OmniBook Ultra 16 (2026) and OmniBook X 14 (2026) will be powered by the new Nvidia RTX Spark superchip. The company expects the HP OmniBook Ultra 16 (2026) and OmniBook X 14 (2026), with the Nvidia RTX Spark chip, to go on sale in select global markets later this year, and the pricing details will be revealed “closer to ...
- India’s AI deal with the UAE challenges U.S. cloud dominance
G42 will deploy U.S.-designed supercomputers in India, offering a new model for governments that want to own their AI hardware.
- The DOJ used Palantir to build an app to help find criminals—and then shut it down
Recently uncovered documents show that the Department of Justice is no longer using a mobile app, built by Palantir, designed to help law enforcement officials search criminal records databases while operating in the field. Thousands of people, including agents at the U.S. Marshals Service, the Drug Enforcement Agency, and the Bureau of Alcohol, Tobacco, Firearms and Explosives, appear to have ended up using the app, which was called “SHIELD”. But by February 2023, the agency had changed its approach and moved to shut it down, according to emails obtained by Fast Company. It’s a reminder that while the government is racing to implement all sorts of law enforcement tech and surveillance technology, these platforms are sometimes rolled back. The beginnings of the app stretch back to at least 2018, when Karl Mathias, then-chief information officer of the U.S. Marshals Service, shared that the agency was in the “alpha” stage of developing a system that would allow officers to ping a crime database run by the FBI and help apprehend people. By the following year, the Marshals had brought on Palantir to help develop the platform. The app was designed to scan a picture of someone’s driver’s license and then run that image in various databases. The effort represented the U.S. Marshals’ new approach to “modernization,” according to comments Mathias made to the trade outlet GovCIO . Since then, the agency said little about Shield, but it seems that the app was ultimately used throughout several law enforcement agencies. One Justice Department IT manager touts on LinkedIn that he helped “develop, deploy, and perform continuous improvement for SHIELD” and that the app enabled more than 2,000 people to perform “real-time biometric and identity verification in the field.” The app was successful enough that the agency requested additional funding to support the platform before fiscal year 2021 and then in fiscal year 2023. Still, emails show that between February and April 2023, technology leaders at DEA, ATF, and Marshals moved to decommission the app, warning the agencies to export and retain any remaining data and shutting down user access when the “end of contract” approached. “User access will be turned off at the end of the contract. Any service accounts we used to communicate with external systems should be identified and disabled,” a DOJ IT manager warned in February 2023. “All cleanup activities should be complete and contractor access and clearances may be closed out.” The Justice Department declined to comment, and the U.S. Marshals Service, the DEA, and the ATF did not respond to repeated requests for comment. Still, one former Palantir employee said that it’s possible the contract itself may not have been particularly lucrative for the company, especially if it was sold to the government as a perpetual license, which would have allowed the Justice Department to use the service indefinitely. A person familiar with the matter says that Palantir has since moved away from the perpetual license model, which led to a transition off partnerships leveraging that model.
- A nurse stole fentanyl and AI missed it, state records say
Sentri7, drug diversion software powered by artificial intelligence and used at hundreds of U.S. hospitals, did not catch a monthslong string of fentanyl thefts in Tennessee in 2025, according to a state document.
- Advancing AI Infrastructure for Agentic AI with NVIDIA DOCA In-Silicon Security
The AI era is driving a new class of infrastructure: AI factories that transform data into intelligence for autonomous AI agents operating at unprecedented...
- Anthropic to offer EU access to its advanced Mythos model
Anthropic is sharing access to its most advanced AI model, Mythos, with the EU after the bloc sought permission over cybersecurity concerns.
Score: 37🌐 MovesJun 1, 2026https://www.cnbc.com/2026/06/01/anthropic-eu-ai-mythos-access-advanced-model.html - Turnitin Adds Customizable AI Assistance to Support Different Assignments, Grade Levels
Turnitin has introduced new customizable settings Turnitin Clarity's built-in AI assistant, enabling instructors to specify AI's role and response complexity for each assignment.
- Wells Fargo CEO: AI’s effect on employment is ‘complicated’
The bank’s biggest AI-related challenge is determining how the technology can transform its business model, Charlie Scharf said Wednesday.
Score: 36🌐 MovesJun 1, 2026https://www.hrdive.com/news/wells-fargo-ceo-scharf-ai-employment-banking-jobs/821585/ - Strava blames zero-code AI apps and scrapers as it tightens API access
The popular fitness-tracking platform, Strava, is restricting access to its API as part of efforts to clamp down on AI scraping, as reported earlier by TechCrunch. Developers who want to build an app using Strava's data now need to pay for a flat $11.99 / month subscription. In an update on its developer hub, Strava […]
Score: 36🌐 MovesJun 1, 2026https://www.theverge.com/gadgets/940854/strava-restricts-api-access-ai-apps - AI-powered finance startup Gradient Labs doubles Series A to $26m
AI-powered finance startup Gradient Labs doubles Series A to $26m
- Adobe Photoshop and Premiere Just Got 2x Faster on Nvidia's New PCs — Creators, Take Note
Adobe Photoshop and Premiere Just Got 2x Faster on Nvidia's New PCs — Creators, Take Note PCMag Middle East
- Singapore's AI registry for public officers
Singapore's AI registry for public officers The Straits Times
- 😹 DuckDuckGo installs up 30% after Google's AI overhaul
PLUS: Anthropic hits $965B, OpenAI wants robots
Score: 36🌐 MovesJun 1, 2026https://www.theneurondaily.com/p/duckduckgo-installs-up-30-after-google-s-ai-overhaul - China has approved the world’s first invasive brain-computer chip—here’s what’s next
One day last October, sitting in the courtyard of his house in China’s Henan province, Dong Hui decided to see if he could hold a pen to write. Dong, 39, had sustained spinal cord injuries in a car accident six years earlier that left him paralyzed from the neck down. Slowly but determinedly, he wrote…
Score: 36🌐 MovesJun 1, 2026https://www.technologyreview.com/2026/06/01/1138133/china-world-first-brain-chip/ - Japan's Nikkei tops 67,000 for first time on AI boost; SoftBank becomes Japan's most valuable firm
Japan's Nikkei tops 67,000 for first time on AI boost; SoftBank becomes Japan's most valuable firm The Straits Times
- How People Are Really Using AI in 2026
One new risk became more salient: letting AI think for you.
- New Mexico Education Committee Recommends AI Oversight Body
While the New Mexico Public Education Department issued optional guidelines on AI use in schools last year, a report given to the Legislative Education Study Committee recommends more oversight and rules.
Score: 35🌐 MovesJun 1, 2026https://www.govtech.com/education/k-12/new-mexico-education-committee-recommends-ai-oversight-body - Google Eases Gemini Usage Limits After Complaints
Google adjusted Gemini usage limits after complaints, adding Pro quota caps, free Flash-Lite prompts, and clearer usage reporting. The post Google Eases Gemini Usage Limits After Complaints appeared first on TechRepublic .
Score: 35🌐 MovesJun 1, 2026https://www.techrepublic.com/article/news-google-gemini-usage-limits-complaints/ - Scaling AI in Health Systems: From Innovation to Sustainable Implementation
Scaling AI in Health Systems: From Innovation to Sustainable Implementation MedCity News
Score: 35🌐 MovesJun 1, 2026https://medcitynews.com/2026/06/scaling-ai-in-health-systems-from-innovation-to-sustainable-implementation/ - The answer to Arizona data center power demands could be on your roof
The answer to Arizona data center power demands could be on your roof azcentral.com and The Arizona Republic
- From pilots to platforms: The next chapter of AI in retail operations
The path toward better serving customers starts with store operations.
- Pat Gelsinger dishes on chips, Intel and Trump
Pat Gelsinger dishes on chips, Intel and Trump
- Erin Brockovich Is Taking on AI Data Centers, and 1 Word Keeps Surfacing in Her Investigation
The famous consumer advocate has launched a new website about data centers. She’s getting flooded with responses.
- Draft federal AI strategy aims to scale up adoption, offer literacy training by 2031
Draft federal AI strategy aims to scale up adoption, offer literacy training by 2031 CBC
Score: 35🌐 MovesJun 1, 2026https://www.cbc.ca/news/politics/federal-artificial-intelligence-strategy-9.7216576 - State of the CIO, 2026: CIOs set the course for AI ROI
Drowning in hype and under pressure from top leadership, CIOs are racing to operationalize strategic AI initiatives in an effort to demonstrate — and more importantly, deliver — measurable ROI from this equally disruptive and transformative technology. The perpetual pipeline of AI pilots and rampant experimentation are giving way to a new mandate to prioritize and scale AI solutions with the greatest propensity to deliver business value and impact the bottom line. Yet the ability to achieve quantifiable ROI from AI projects remains somewhat elusive for many CIOs as they spearhead AI’s next chapter. According to CIO.com’s 25th annual State of the CIO survey , which canvassed 662 IT leaders and 249 line of business users, less than a fifth (19%) of respondents say AI initiatives have met or exceeded business goals, and 18% admit fewer than a third of AI use cases are meeting defined expectations. CIO.com / Foundry Lack of clarity around business strategy and metrics is proving to be a major barrier to advancing CIOs’ AI agendas. Almost a third (32%) of this year’s State of the CIO respondents called out ill-defined ROI metrics as a hurdle to scaling AI, along with murky corporate AI strategy (31%) and a lack of in-house expertise , cited by 40% of this year’s respondents. “No one is measuring ROI on an ongoing basis because we are facing counterpressures from every vice president and line-of-business domain looking to implement AI for their own optimization,” notes Andrea Ballinger , CIO at Rensselaer Polytechnic Institute (RPI). “We are saying yes to everyone without stepping back and focusing on the business cases that show real value.” CIO.com / Foundry Creating the infrastructure to achieve AI ROI The dynamic is starting to shift as organizations prioritize targeted use cases and lay the groundwork for scalability and ROI. Cross-functional steering committees and specialized task forces are emerging as critical building blocks to identify, prioritize, and align use cases to enterprise goals. Eighty-three percent of IT leaders surveyed confirmed their organizations either have such structures in place or are planning to implement them within the year. IT is the dominant player on these committees with other functional areas well represented, including corporate leadership and security and risk teams, and to a lesser degree, business-oriented domains such as finance, legal, and HR. Formal processes for approving AI projects are far less evolved. Slightly more than half (53%) of 2026 State of the CIO respondents have established some type of official approval process, with 28% planning to activate something within the next 12 months. KPIs, another critical milestone for AI success , are also not well defined in most enterprises. Less than half (47%) of respondents have established formal metrics, with another 34% planning to do so within the year. For those measuring AI success, operational efficiency and process improvement rank as the top metric, cited by 40% of respondents, followed by employee productivity (34%) and cost reduction (30%). AI’s impact on revenue or growth is less of a factor today, cited by only 27% of respondents. CIO.com / Foundry First Student, a leading provider of school bus transportation services, has stood up a well-defined innovation framework and AI-specific council — two moves CIO Sean McCormack credits to the firm’s early success scaling AI initiatives aligned to key business goals. The AI council, with representation across business leaders and the C-suite, meets regularly to review AI use cases and identify those with most potential for payback. “We have more discipline around business cases than most companies,” says McCormack. “Everything is metrics-driven and dependent on proving value. By the time we put something into production, it’s been through a series of proof of concepts, there’s been a deep dive on financials, and we are able to move quickly and demonstrate value.” Three years into its AI journey, TIAA has a rich stable of generative AI and agentic AI use cases spanning fraud detection and prevention along with call center companions and a litany of other tools. The majority (85%) of the financial services firm’s workforce uses TIAA Gate, its internal AI platform. Yet even with all the right structures in place — investment in training, robust governance frameworks, steering committees, an AI center of excellence (CoE), and an enterprise mandate for strategic use of AI as part of everyone’s performance goals — ROI remains a challenge. “What’s on paper sometimes doesn’t turn into real ROI given the reality of operational costs,” notes Sastry Durvasula , TIAA’s chief operating, information & digital officer. “Something may prove to be a successful pilot, but you need to understand the full cost of operations — for example, the efficiencies of running tokens or how you’re handling traffic or RAG [ retrieval augmented generation ].” Thomas Prommer, a longtime CTO, CIO, and CAIO, makes three recommendations to facilitate the quest for AI ROI. First, establish joint accountability at the project level with a named technical and business sponsor for each project, both of whom co-own outcomes. His firm had success replacing a centralized AI CoE with embedded AI squads that live inside individual business units. The CoE model created a clearinghouse that nobody owned, whereas embedded teams force accountability at the point of business impact, he explains. Prommer’s third recommendation is to implement stage-gated funding tied to outcome milestones, not deliverable milestones. “We don’t fund ‘build a model,’ we fund ‘reduce returns by 8% on this category’ with checkpoints at 90, 180, and 270 days,” he explains. “Projects that miss two checkpoints gets killed. We kill roughly a third of what we start and that’s healthy.” Developing a keen understanding of business workflows and engineering the experience layer for the people tasked with executing AI-enriched workflows is essential to creating value and effective adoption at scale. “If someone on the data science team builds a great model that provides insights on improving manufacturing efficiency, but it’s so far removed from what the shop floor supervisor does in day-to-day life, it will never be used at scale,” says Sriram Krishnasamy , the former chief digital information and transformation officer at FedEx. The CIO as chief AI orchestrator Who better than the CIO to usher in the organizational structures and business practices that help identify the right AI use cases and establish metrics for success. IT leaders’ in-depth knowledge of AI and the broader technology stack is a plus. But the reason top executives are leaning on CIOs as critical orchestrators of AI is their proven ability to work effectively across business functions and serve as change management champions . Almost half (46%) of this year’s respondents view the CIO as a business leader who proactively identifies business needs and opportunities and follows up with technology and provider recommendations that align with stated business goals. The vast majority (83%) view CIOs as a changemaker. Much like last year, the top CEO priority for IT leaders is to research and implement AI products and projects, cited by 27% of respondents. IT leaders are meeting the mandate by working far more closely with lines of business on AI applications, according to 79% of respondents. “AI requires so much executive engagement — in our case, it made the most sense for me to lead the charge,” says First Student’s McCormack, also a member of CIO.com’s Hall of Fame . This year’s IT leader respondents expect to accelerate and expand involvement with AI/machine learning (76%) and agentic AI (70%), as well as cybersecurity (63%). Responding organizations anticipate a boost in investments across the full complement of AI technologies, including generative AI (67%), machine learning (66%), and agentic AI (65%) over the coming year. CIO.com / Foundry First Student is currently running AI in production at scale across multiple use cases, including for predictive maintenance, fleet and driver safety, contract development, automated hiring, agentic software development, and agentic voice bots that assist internal users with help desk and HR issues. As the company builds out its portfolio of AI-enabled use cases, McCormack says it’s critical to ground everything in a flexible architecture. “It’s such a changing landscape; it’s difficult to pick a solution,” he explains. “We’re building our own architecture so we can quickly switch models” and not be dependent on one system, he says. The rest of the IT agenda While AI commands an all-hands-on-deck approach, it’s not the only CEO directive for CIOs this year. Cyber and data security remain top-of-mind in the C-suite, with a quarter of State of the CIO respondents noting it as a top CEO priority for 2026, up from 20% last year. CEOs are also looking for CIOs to strengthen IT and business collaboration, cited by 23% of 2026 respondents. To achieve those directives, CIOs are expanding technology initiatives in areas such as business process and IT automation (56%), security and risk management (55%), and data and business analytics (54%). CIO.com / Foundry Building a solid data and governance foundation is the most important agenda item at RPI, primarily in preparation for more extensive AI deployment and adoption. Ballinger, 70 days into her tenure as RPI CIO, concedes the educational institution is not aiming to be on the first adopter wave of AI innovation. Rather, its strategy is to adapt, grow, and transform organizational structures and its data ecosystem in pursuit of maximizing AI’s promised advantages. Ballinger is currently shopping an RFP that encompasses a data fabric layer, secure containerization, and a data factory approach. “We are designing the ecosystem with strategy and KPIs in mind,” she says. “The entire process is predicated on deciding what business cases are valuable before they hit the data ecosystem. We’re not looking to put something into the ecosystem and hope it gets a return.” With the pace of change growing more intense and the stakes surrounding AI innovation soaring higher, the CIO role continues to be more business-oriented and less straight technology focused. Eighty-four percent of this year’s IT leader respondent pegged the CIO position as more digital and innovation focused while 82% confirmed that CIOs are more likely to actively lead digital transformation efforts compared to their business counterparts. With the average CIO now juggling 1.6 positions, including chief security officer, CISO, CAIO, and other business-related posts, the job continues to become more expansive, highly strategic, and more fulfilling, especially for leaders willing to close the door on the traditional “keeping the lights on” CIO model and embrace new challenges. “The CIO of 2026 is a hybrid — half operating architect, half risk officer,” Prommer says. “The technology choices are getting easier, but the business and ethical choices are getting harder. CIOs who only know the tech stack will be reporting to CIOs who know both.”
Score: 35🌐 MovesJun 1, 2026https://www.cio.com/article/4178006/state-of-the-cio-2026-cios-set-the-course-for-ai-roi.html - AI-led weather forecasting systems for Africa
AI-led weather forecasting systems for Africa physics.ox.ac.uk
Score: 35🌐 MovesJun 1, 2026https://www.physics.ox.ac.uk/news/ai-led-weather-forecasting-systems-africa - Cars24 launches AI lab, commits $20 million to AI, early-stage startups
Cars24 is launching Cars24 Labs, a new initiative backed by a $20 million investment. This program will focus on developing artificial intelligence products and supporting early-stage founders. Cars24 aims to provide capital, mentorship, and infrastructure to help innovators build the next generation of AI businesses. The company is also preparing for a public listing.
- Anthropic is giving the E.U.'s cybersecurity agency access to its restricted AI security model
ENISA will join Project Glasswing, Anthropic's program for testing the powerful vulnerability-finding model before a wider release
- Anthropic’s Mythos Is a Security Powerhouse. It’s Also a Budget Buster
Anthropic’s Mythos Is a Security Powerhouse. It’s Also a Budget Buster The Information
Score: 35🌐 MovesJun 1, 2026https://www.theinformation.com/articles/anthropics-mythos-security-powerhouse-budget-buster - Nvidia Has Best Trading Day In Months After Revealing New Microsoft Chip
The new line of Windows laptops and desktops will be the first ever to run solely on an Nvidia-made chip, the RTX Spark, as the AI giant looks to take on rivals Intel and AMD.
- Merge launches Agent Handler for Employees as an IT gatekeeper for workplace AI agents
Merge API Inc., a platform provider delivering connective infrastructure for artificial intelligence to business data and tools, launched Agent Handler for Employees, easing the strain on information technology teams. With today’s release, Merge is offering a product that connects to identity providers, imports employees and groups, maps them to approved tools and actions across AI […] The post Merge launches Agent Handler for Employees as an IT gatekeeper for workplace AI agents appeared first on SiliconANGLE .
Score: 35🌐 MovesJun 1, 2026https://siliconangle.com/2026/06/01/merge-launches-agent-handler-employees-gatekeeper-workplace-ai-agents/ - Stargate live updates: OpenAI's Altman says 'people are right to be anxious' about AI
OpenAI, Oracle and Related Digital discuss the Stargate Michigan buildout, AI infrastructure, jobs and community plans.
Score: 35🌐 MovesJun 1, 2026https://www.cnbc.com/2026/06/01/stargate-project-michigan-live-updates.html - AI costs how much? GitHub Copilot users react to new usage-based pricing system.
Some report burning through their whole monthly "AI credit" allotment in a single day.
- NVIDIA DSX OS Delivers Open, Modular Software for Operating AI Factories at Scale
AI is now essential infrastructure, powered by AI factories that generate intelligence in the form of tokens. As demand grows, these factories must scale...
- AI was supposed to prevent downtime. Instead, it’s creating new kinds of outages
Enterprise AI promised executives something close to operational certainty: fewer outages, less human error, and systems capable of catching problems before customers ever noticed. But a new report from the software company Splunk on AI-related downtime suggests those promises are colliding with a messier reality. For businesses, downtime—unexpected interruptions to the software systems and applications that keep operations running—can trigger everything from lost sales to frozen logistics networks and customer backlash. For years, companies treated the problem as fundamentally solvable: Automate enough of the right processes, and human error could largely be engineered out. Acting on that logic, companies spent a median of $24.5 million annually on artificial intelligence systems designed to prevent downtime, per the report from Splunk, a unit of Cisco. But many now report that AI itself is becoming part of the outage problem, quietly introducing several new failure modes in the process. Half of surveyed organizations experienced downtime tied to incorrect AI automation or model drift. Nearly one-third blamed bugs introduced by embedding AI into production systems. Conducted with Oxford Economics across 2,000 executives of Global 2000 companies, the survey report estimates that unplanned downtime now costs businesses $600 billion annually, up 50% in just two years. Every minute of downtime costs roughly $15,000, and businesses lose an average of $300 million annually before anyone formally calls it a crisis. Splunk calls it the reliability paradox: The more aggressively companies deploy AI to eliminate operational risk, the more they find themselves managing a newer, less predictable category of it. “Organizations are deploying AI into mission-critical systems without clearly defined escalation paths,” says Kamal Hathi, senior VP and general manager of Splunk. “They lack monitoring tuned to detect model drift, and there’s no clear ownership when things go wrong.” The financial exposure extends far beyond IT budgets. Hathi notes that stock prices drop an average of 3.4% per major incident, ransomware payouts have nearly tripled to $40 million, and regulatory fines now average $51 million. AI Built to Reduce Risk Is Now Manufacturing It The AI race rewards speed above almost everything else. What began with copilots and chat interfaces is accelerating toward autonomous agents, often without a human in the loop. That velocity is also changing what failure looks like. Hathi says companies are not misreading AI’s value so much as underestimating what responsible deployment requires. There is a tendency to treat AI deployment like a software upgrade. But AI learns from shifting environments and interacts with systems in ways that do not follow deterministic logic. “Resilience can’t be an afterthought,” he says, referring to the ability to absorb disruption, recover quickly, and maintain continuity. The report found that 44% of organizations use agentic AI, yet 68% worry those systems may behave unpredictably. The range of types of attack is widening as well. Prompt injection and data poisoning, two forms of AI-targeted attacks in which bad actors manipulate what an AI system sees or learns to alter its behavior, are on the rise. Nearly one in four organizations has encountered them, and 77% of technology leaders believe cybercriminals armed with generative AI will increase downtime at their organizations. “Agentic systems need to earn their autonomy incrementally,” Hathi says. “They must be governed by visibility and accountability at every step — not deployed at scale and monitored retroactively.” The Silent Failure Mode Nobody Planned For Greg Leffler, director of developer evangelism and lead evangelist at Splunk, says AI-related downtime rarely resembles a traditional outage. Instead of a dramatic collapse, it often looks like a compounding erosion of system behavior that spreads long before anyone thinks to investigate it. He pointed to two patterns appearing repeatedly across enterprise environments. The first is model drift, which he describes as “an automation pipeline making correct decisions six months ago whose training data no longer reflects current traffic. By the time anyone notices, the damage is already spreading across interconnected services.” The second is broken integrations, where an AI system acts on incomplete data and triggers a chain of failures across connected systems that no single team fully owns or monitors end to end. Both degrade confidence gradually, until something critical finally tips over. AI systems are too often deployed with the assumption that they are self correcting, an assumption traditional infrastructure was never allowed to make. “The engineering discipline applied to software releases—staged rollouts, canary testing, rollback procedures—must now apply to every production model carrying decision-making authority,” Leffler says. The report’s sharpest finding, however, is not about model capability but about who is in control. Only 38% of surveyed technology executives reported consistently identifying the root cause of downtime incidents, despite heavy investment in monitoring platforms. Leffler explained that as automation absorbs more routine operational decisions, fewer engineers develop the deep system intuition needed to diagnose failures when automation breaks. At the same time, today’s tech stacks rely heavily on external AI providers and third-party services that teams have little direct visibility into, creating what he calls a compounding opacity problem: layers of interconnected risk sitting largely outside what can be observed. “Agentic systems should independently diagnose issues, execute routine fixes, and perform code rollbacks—but escalate any higher-stakes decision for human approval,” Leffler says. He adds that the challenge is as much cultural as technical. “If engineering teams aren’t measuring reliability with the same rigor they measure velocity, governance frameworks will always lose to ship timelines.” Shadow AI Is Outpacing Enterprise Visibility Some of the hardest problems to quantify, and perhaps the hardest to fix, are happening outside the official technology stack. Earlier generations of “shadow IT” typically involved employees adopting unapproved software, cloud services, or collaboration tools outside formal IT oversight, creating security and compliance headaches. Shadow AI raises the stakes. Fully 66% of organizations report employees using unapproved AI tools at work to write code, generate business outputs, and automate decisions, often without centralized visibility into what data those tools access or how their recommendations influence production environments. Unlike shadow IT, shadow AI can shape operational behavior while leaving little trace of how or why decisions were made. “It’s all three: a policy problem, a visibility problem, and a governance problem,” Hathi says. “Policy alone won’t solve it. Organizations need to deploy an evaluation system for what AI should do, backed by a telemetry layer grounded in logs, metrics, and traces.” AI will keep getting smarter. The harder challenge, and the one most enterprises are only beginning to confront, is building systems capable of seeing and correcting intelligent behavior before it becomes a business crisis. “Every competitor now has access to similar models and cloud infrastructure,” Hathi says. “Resilience, governance, and observability are becoming the real differentiators. The enterprises that internalize that first will define what operational excellence means in the AI era.”
- Human Brain Cells Grown on a Chip Level Up to Play 'Doom'
“We are just scratching the surface of what these neural cultures can achieve.” ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
Score: 35🌐 MovesJun 1, 2026https://www.sciencealert.com/human-brain-cells-grown-on-a-chip-level-up-to-play-doom - Miro’s Big Bet: Can A Whiteboard Company Become The AI Decisioning Layer For The Enterprise?
At its Canvas 26 conference, Miro claimed that it is no longer in the whiteboard business. Instead, it wants to become the collaborative decision-making layer for the agentic enterprise. That’s bold and ambitious repositioning and, potentially, timely. As the cost of intelligence is plummeting, the amount of work organizations can generate is increasing significantly — […]
- Microsoft Ends Claude Code Licenses As It Shifts Developers To Copilot
Microsoft ends Claude Code licenses and shifts developers to its in‑house Copilot model, signaling a strategic move toward AI self‑sufficiency and distribution power.
- OpenAI’s next legal battle is against states that claim its models are dangerous
Florida Attorney General James Uthmeier sued OpenAI for putting children at risk, and said other states will likely also launch legal actions.
- Nvidia courts Korea's tech giants at Taipei dinner as AI push deepens
Nvidia courts Korea's tech giants at Taipei dinner as AI push deepens Reuters
- Tesla Stock Drops. OpenAI Is Coming for One Line of Its Business.
Tesla Stock Drops. OpenAI Is Coming for One Line of Its Business. Barron's
- OpenAI enhances ChatGPT app with this hidden feature
OpenAI recently added a new feature to the ChatGPT app: a hidden gesture that lets you instantly set the effort level of a prompt. ChatGPT also introduced a new way to easily navigate longer conversations. Meanwhile, the default model used by ChatGPT has changed since arriving last month. more…
Score: 34🌐 MovesJun 1, 2026https://9to5mac.com/2026/06/01/openai-enhances-chatgpt-app-with-this-hidden-feature/ - Amazon is sharing its AI shopping technology with other retailers through AWS
The AWS Agentic Shopping Assistant packages the architecture behind Alexa for Shopping, and Kate Spade has already deployed it
- Iranian hackers are weaponising ChatGPT and Gemini for cyberattacks on the US and Israel: Report
Iranian hackers are leveraging AI models like ChatGPT and Gemini to enhance their cyber operations against the US and Israel
- UAE launches AI leadership programme to accelerate National AI Strategy 2031
UAE launches AI leadership programme to accelerate National AI Strategy 2031 Arabian Business