AI News Archive: May 27, 2026 — Part 2
Sourced from 500+ daily AI sources, scored by relevance.
- Qualcomm Signs AI Chip Deal with ByteDance to Power Data Center AI Agents
Qualcomm to supply millions of ASIC chips to ByteDance data centers for AI agent workloads, as the chipmaker diversifies beyond mobile into cloud infrastructure.
- China Tightens Online Pharmacy Rules on AI Use and Prescriptions
The country aims to curb abuses in obtaining prescription medication online — a practice that has become easier with the rise of pharmaceutical e-commerce platforms.
Score: 60🌐 MovesMay 27, 2026https://www.sixthtone.com/news/1018580/China Tightens Online Pharmacy Rules on AI Use and Prescriptions - NVIDIA Blackwell Sets STAC-AI Record for LLM Inference in Finance
Large language models (LLMs) are revolutionizing the financial trading landscape by enabling sophisticated analysis of vast amounts of unstructured data to...
Score: 60🌐 MovesMay 27, 2026https://developer.nvidia.com/blog/nvidia-blackwell-sets-stac-ai-record-for-llm-inference-in-finance/ - Cisco and OpenAI redefine enterprise engineering with Codex
Cisco and OpenAI are redefining enterprise engineering with Codex, helping Cisco scale AI-native development, accelerate AI Defense work, and automate defect remediation.
- MR-AIV reveals in vivo brain-wide fluid flow with physics-informed AI
Science Advances, Volume 12, Issue 22, May 2026.
- Zuckerberg's philanthropic venture unveils AI world model for drug discovery
Zuckerberg's philanthropic venture unveils AI world model for drug discovery Reuters
- An All AI-Generated Film Is Coming to Tribeca and It’s Sure to Raise Eyebrows
Dreams of Violets , made over three months, dramatizes the civil unrest in Tehran from this past January
Score: 60🌐 MovesMay 27, 2026https://www.rollingstone.com/tv-movies/tv-movie-news/dreams-of-violets-ai-movie-tribeca-1235569122/ - Meta Launches Meta One: A Strategy to Monetize AI for Instagram, Facebook Users
Meta Launches Meta One: A Strategy to Monetize AI for Instagram, Facebook Users Barron's
- Wayve CEO Alex Kendall on the AI Revolution in Autonomous Driving
Wayve CEO Alex Kendall on the AI Revolution in Autonomous Driving Automotive News
Score: 60🌐 MovesMay 27, 2026https://www.autonews.com/video/wayve-ceo-alex-kendall-on-the-ai-revolution-in-autonomous-driving/ - Inside China’s Push to Build an Army of AI-Powered Combat Robots
China’s missile-armed robot dogs and humanoid systems point to a new phase in military robotics, raising questions about AI warfare. The post Inside China’s Push to Build an Army of AI-Powered Combat Robots appeared first on TechRepublic .
Score: 60🌐 MovesMay 27, 2026https://www.techrepublic.com/article/news-china-military-robotics-ai-warfare-apac/ - AI costs begin to bite as agents may increase token demand by 24 times, says Goldman Sachs report — Uber and Microsoft among companies feeling the bite of tokenized billing
Major tech companies are considering refining their approaches to AI, as rising token costs and increased token demand from AI agents make the costs harder to justify, with limited return on the investment.
- Uber’s president doesn’t know what the company’s 10.5 million drivers will be doing in 15 years
Andrew Macdonald tells Toronto Tech Week’s Homecoming that the gig app’s labour pool may soon start to shrink. The post Uber’s president doesn’t know what the company’s 10.5 million drivers will be doing in 15 years first appeared on BetaKit .
- Google Moves AI Agents into the Mainstream
At its recent I/O developer conference, Google presented artificial intelligence agents not as a distant research project, but as a product strategy spanning Search, personal assistants, productivity software, developer tools, and smart glasses.
Score: 59🌐 MovesMay 27, 2026https://campustechnology.com/articles/2026/05/27/google-moves-ai-agents-into-the-mainstream.aspx - ECB tells banks to invest more to get a grip on AI security risk
ECB tells banks to invest more to get a grip on AI security risk Reuters
- Authors Sue Meta's AI Scientists Directly in Llama Copyright Case
A proposed class action targets not just Meta and Mark Zuckerberg but the research scientists who allegedly carried out the company's mass piracy of tens of millions of books.
- Using AI to advance brain therapeutics
Using AI to advance brain therapeutics EurekAlert!
- Salesforce shares sink on soft revenue outlook, as AI disruption concerns linger
Salesforce shares sink on soft revenue outlook, as AI disruption concerns linger
Score: 59🌐 MovesMay 27, 2026https://www.marketwatch.com/bulletins/redirect/go?g=0bf347ac-608d-493f-9e4a-65a79bb0c5ef&mod=mw_rss_bulletins - The AI GPU market has belonged to hyperscalers, but AMD’s MI350P is coming for the enterprise
Enterprise GPU access is expanding as Advanced Micro Devices Inc. lowers barriers with the air-cooled Instinct MI350P GPU and a ready-to-run software stack designed for standard enterprise servers. The next wave of enterprise GPU adoption hinges not just on raw silicon performance, but on dramatically lowering the barrier to deployment. AMD is betting that pairing […] The post The AI GPU market has belonged to hyperscalers, but AMD’s MI350P is coming for the enterprise appeared first on SiliconANGLE .
Score: 58🌐 MovesMay 27, 2026https://siliconangle.com/2026/05/27/enterprise-gpu-access-mainstream-data-centers-delltechworld/ - Bank of America sharply lifts Korea outlook on AI chip boom
Bank of America has sharply raised its 2026 growth forecast for South Korea, citing an artificial intelligence-driven semiconductor boom that it said is strong enough to lift exports, investment, consumption and government revenue. The bank upgraded its 2026 GDP growth forecast for Korea to 3.1 percent from its previous estimate of 1.9 percent, calling the revision "a clear shift to above-trend expansion," according to its report released Tuesday. It also raised its current account surplus proje
- NYC insurance AI startup Pace raises $46M just months after previous round
The company announced a $46 million Series B led by Thrive Capital and Sequoia Capital. It raised $10 million in January.
Score: 58💰 MoneyMay 27, 2026https://www.bizjournals.com/newyork/news/2026/05/27/pace-raises-46-million.html?ana=brss_6150 - Uber Says Its AI Costs Just Aren’t Worth It
"AI is not free." The post Uber Says Its AI Costs Just Aren’t Worth It appeared first on Futurism .
Score: 58🌐 MovesMay 27, 2026https://futurism.com/artificial-intelligence/uber-ai-costs-arent-worth-it - AI video analysis startup Airis Labs raises $60M
Airis Labs Ltd., a provider of video analysis software for government agencies, launched today with $60 million in funding. The startup raised just over half the capital through a Series A round led by PSG Equity. The growth equity firm was joined by TLV Partners, Stepstone Group, Redseed Ventures and multiple angel investors. Airis provides […] The post AI video analysis startup Airis Labs raises $60M appeared first on SiliconANGLE .
Score: 58💰 MoneyMay 27, 2026https://siliconangle.com/2026/05/27/ai-video-analysis-startup-airis-labs-raises-60m/ - Public PE titans are saying goodbye to software, hello to AI
Public PE titans are saying goodbye to software, hello to AI PitchBook
Score: 58🌐 MovesMay 27, 2026https://pitchbook.com/news/articles/public-pe-titans-are-saying-goodbye-to-software-hello-to-ai - Jan Leike Joins Anthropic: What It Actually Meant
The quiet defection that changed the terms of the safety debate. Jan Leike, co-lead of OpenAI’s superalignment team, the researcher who spent three years trying to build a scientific framework for controlling AI systems smarter than humans, joined Anthropic on May 28th, 2024. He did not go quietly. Two weeks before he announced his new role, he posted a resignation letter on X that read, in part: “Safety culture and processes have taken a back seat to shiny products.” OpenAI dissolved the superalignment team. Leike packed the mission into a cardboard box and walked it across town to Dario Amodei. TechCrunch reported that he would report directly to Jared Kaplan, Anthropic’s chief science officer, and lead a new team working on scalable oversight, weak-to-strong generalization, and automated alignment research. In practice, Anthropic had hired the person who wrote OpenAI’s original case for taking superintelligence alignment seriously, gave him a direct line to the scientific leadership, and pointed him at the hardest open problem in AI safety: how do you control a model that’s smarter than the people trying to control it? His decision to join Anthropic came as a surprise to most. First, he had built his career at the labs most likely to produce what he was trying to align. He spent years at DeepMind, then joined OpenAI in 2021. He co-led the superalignment project with Ilya Sutskever, which OpenAI announced in June 2023 with a promise of 20% of the company’s compute and a four-year deadline to solve alignment for superintelligent systems. Leaving that — even after the team was dissolved — meant giving up institutional access to the very models he was trying to study. Second, he had spent years arguing, implicitly, that the best place to do safety work was inside the labs building the dangerous systems. The logic was practical: you can’t align what you can’t see. External researchers work with limited access and lagging information. The alignment team at OpenAI had first-mover access to every new capability jump, sometimes before the research papers were written. Joining Anthropic meant accepting a form of institutional distance from the systems he was most worried about GPT-5 and whatever came after it. Third, he had not shown any obvious signs of wanting to leave. His departure, when it came, was abrupt enough that observers initially assumed something had gone catastrophically wrong internally. The actual cause was more specific and, once known, more alarming: OpenAI had decided that the superalignment work could be absorbed into the broader research organization without a dedicated team. The four-year deadline to solve superintelligence alignment became a footnote. These three conditions created a researcher who was suddenly available and, more importantly, suddenly motivated. The question was where someone with his profile would go. The answer told you something about where the field was actually heading. In his resignation post, Leike was more pointed than researchers of his stature usually allow themselves to be in public. He wrote that OpenAI was not adequately prioritizing safety, that the company had created a culture where safety concerns were treated as a tax on progress rather than a core constraint. He was careful not to accuse anyone of acting in bad faith. He did not have to. The structure of his criticism — the gap between the public commitments and the internal resource allocation — made the point without needing to name anyone. When he announced his move to Anthropic three weeks later, the framing was not about escape. It was about continuity. He called it “continuing the superalignment mission.” The word choice mattered. He was not starting a new project. He was relocating an existing one. [X post — Jan Leike, May 28, 2024: “I’m excited to join @AnthropicAI to continue the superalignment mission! My new team will work on scalable oversight, weak-to-strong generalization, and automated alignment research. If you’re interested in joining, my DMs are open.”] The three areas he named — scalable oversight, weak-to-strong generalization, automated alignment research — are technical shorthand for a single underlying problem. As models become more capable, the humans overseeing them become relatively less capable of evaluating whether what the model is doing is actually good. Scalable oversight is the research program for solving that evaluation gap. Weak-to-strong generalization is the empirical question of whether a weaker model can meaningfully guide a stronger one. Automated alignment research is the recursive version of both: can AI systems help us figure out how to align AI systems? This was not the kind of safety work that involved writing policy documents or publishing red-teaming reports. It was the kind that required access to frontier models, significant compute, and a team willing to work on problems with no clear solution timeline. Anthropic offered all three. But why had he chosen Anthropic now? He had other options. DeepMind had a safety program. He could have started something independent. The academic research community would have welcomed him. The answer, visible only in retrospect, is that Leike had watched the internal dynamics at OpenAI long enough to form a specific theory about what kind of institution could actually do this work. Not a lab that treated safety as a constraint to be satisfied at release time. Not a university group without the model access. An institution where safety research sat adjacent to capability research, with enough organizational weight that it couldn’t be quietly absorbed when it became inconvenient. Anthropic was founded by people who had left OpenAI over exactly this disagreement. Dario Amodei, Daniela Amodei, and the original team departed in 2021 because they believed that the commercial trajectory of a sufficiently successful AI company would eventually crowd out safety considerations. They built Anthropic on the premise that the only way to take safety seriously was to make it structurally central, not a team that existed at the pleasure of the product roadmap. Leike had watched that thesis stress-tested from the outside for three years. He had seen what happened when a safety team’s compute allocation came up for review. He had seen what happened when a four-year deadline met a product launch cycle. He joined Anthropic because the thesis had held, and because the alternative — continuing to believe that safety work could survive intact inside a lab optimizing for something else — had demonstrably failed. [X post — Jan Leike, May 4, 2024, resignation post: “Stepping away from this job has been one of the hardest things I have ever done, because we urgently need to figure out how to steer and control AI systems much smarter than us.”] What he was actually signing up for at Anthropic was considerably more technically demanding than the phrase “AI safety researcher” implies in public discourse. Weak-to-strong generalization is an open empirical question with no known solution. The basic setup: you take a weaker model and use it to supervise a stronger one. The stronger model will, in many cases, perform better than its supervisor on the tasks it’s being evaluated on. The question is whether the supervision is meaningful at all when the supervisor can’t fully evaluate the output. If it isn’t, you have a fundamental problem: as models get more capable, the humans and weaker models overseeing them become systematically less able to catch errors, deceptions, or misaligned objectives. This is not a hypothetical concern for 2040. It is a practical engineering problem for the generation of models being trained right now. Claude’s capability level is already high enough that Anthropic’s own researchers sometimes cannot fully verify whether a given output reflects genuine reasoning or confident pattern-matching. The alignment team Leike was building was, in part, trying to build the tools to answer that question before it became unanswerable. John Schulman, another OpenAI co-founder, was already at Anthropic by the time Leike arrived. Nick Joseph, who would later become Karpathy’s direct manager on the pre-training team, had made the same crossing nine months earlier. The pattern had become legible enough that the online community had started calling it the “OpenAI diaspora” — a one-way flow of senior technical people choosing Anthropic over the alternatives, for reasons that each of them articulated differently but that converged on the same structural assessment. The field was splitting along a fault line that had been forming since 2021. On one side: labs where safety was a function of the product team, responsive to timelines and market pressure. On the other: a lab that had staked its organizational identity on the premise that the two couldn’t be separated. The diaspora was the revealed preference of people who had seen both sides from the inside. Karpathy’s arrival in May 2026, two years after Leike’s, completed something. Not a team — the team had been building the whole time. Something more like a thesis. The capability researchers and the alignment researchers, separately and for their own reasons, had both concluded that Anthropic was where the serious work was going to happen. Leike’s autoresearch equivalent is harder to name. He doesn’t release things publicly the way Karpathy does. What he produces is slower and less visible: interpretability findings, scalable oversight frameworks, alignment evals that will eventually be used to decide whether Claude is safe enough to deploy at the next level of capability. The Karpathy Loop is the informal name for the recursive research cycle that could make Claude better at training itself. There is no catchy name for what Leike is building, but the dependency is directional. You cannot safely deploy a recursively self-improving system without a prior answer to the alignment problem that Leike’s team is working on. That is the picture that Leike could see when he chose Anthropic over the alternatives. Not a company that was winning the benchmark race or dominating the revenue charts. A company that had committed to solving both problems at the same time, in the same building, with the same organizational weight behind each. Whether that bet pays off is still the open question. The model that can accelerate its own pre-training will arrive before anyone has proved that weak-to-strong supervision works at the scale required to oversee it. To us, this is an interesting research problem. To them, it is a countdown. Every time I write about recursive self-improvement, I use this lithograph. It remains the most honest diagram of the problem. Jan Leike Joins Anthropic: What It Actually Meant was originally published in Towards AI on Medium, where people are continuing the conversation by highlighting and responding to this story.
- Meituan Open-Sources LongCat-Video-Avatar 1.5: Photorealistic Digital Human Video Framework
Meituan releases version 1.5 of its open-source digital human video generation framework, achieving state-of-the-art lip-sync accuracy with just 8 inference steps.
- Alibaba’s new AI model scores higher than OpenAI, Google rivals in coding ranking
Alibaba Group Holding’s latest artificial intelligence model has clinched a top-tier spot on a major global coding leaderboard, making the Chinese technology giant the only developer other than Anthropic to break into the ranking’s top five spots. Qwen3.7-Max, Alibaba’s latest AI model, scored 1,541 on the Code Arena ranking to claim the fourth spot globally, placing it ahead of rival models from OpenAI and Google. The other four spots in the top five were held by various iterations of Claude...
- On the possibility of joint U.S.–China AI guidelines
On the possibility of joint U.S.–China AI guidelines The Washington Post
- Fujitsu develops self-evolving multi-AI agent technology that learns and adapts to business operations
Fujitsu Limited today announced the development of a self-evolving multi-AI agent technology (1) that enables multiple AI agents to perform tasks as a team, continuously and safely learning from daily execution results, human feedback, policy revisions, and specification changes. In corporate operations, legal revisions, system changes, specification updates, and on-site rule modifications occur continuously. In […] The post Fujitsu develops self-evolving multi-AI agent technology that learns and adapts to business operations appeared first on CXOToday.com .
- Goldman Sachs says hedge funds are ‘doubling down’ on AI but dumping software stocks
Hedge funds and mutual funds moving toward semiconductors with some shared favorites
Score: 58🌐 MovesMay 27, 2026https://www.cnbc.com/2026/05/25/hedge-mutual-funds-semiconductors-software-goldman-.html - China expands travel curbs to top AI talent at private firms
China expands travel curbs to top AI talent at private firms The Japan Times
Score: 58🌐 MovesMay 27, 2026https://www.japantimes.co.jp/business/2026/05/27/tech/china-travel-curbs-ai-firms/ - Kuaishou’s Kling AI Video Unit Reaches $500 Million in Annualized Revenue
Kuaishou’s Kling AI Video Unit Reaches $500 Million in Annualized Revenue The Information
Score: 58🌐 MovesMay 27, 2026https://www.theinformation.com/briefings/kuaishous-kling-ai-video-unit-reaches-500-million-annualized-revenue - ElevenLabs’ new music-generation model can switch genres mid-track
ElevenLabs' new model will let users regenerate a section of a song without affecting the rest of the track.
Score: 57🤖 ModelsMay 27, 2026https://techcrunch.com/2026/05/27/elevenlabss-new-music-generation-model-can-switch-genres-mid-track/ - AI generates entire chemical formulations for battery electrolytes
AI generates entire chemical formulations for battery electrolytes EurekAlert!
- Snowflake’s stock soars as AI acceleration drives record product-revenue growth
Snowflake’s stock soars as AI acceleration drives record product-revenue growth
Score: 57🌐 MovesMay 27, 2026https://www.marketwatch.com/bulletins/redirect/go?g=50225957-bb9c-440b-a598-4699682f2b56&mod=mw_rss_bulletins - Last Week in AI #341 - Musk loses to OpenAI, Google's IO updates, OpenAI solves Erdős
Elon Musk Loses $150 Billion Suit Against OpenAI and Sam Altman, Google updates its Gemini app to take on ChatGPT and Claude at IO 2026, and more!
- CERT-In’s new AI cybersecurity guidelines call for 12-hour patch windows for critical flaws: Key takeaways
CERT-In’s new AI cybersecurity guidelines call for 12-hour patch windows for critical flaws: Key takeaways
- Preventing a ‘Chernobyl moment’ in AI
A White House order on testing frontier models would be a significant first step
- Amazon's New AI Creators' Fund Sees Prime Video Greenlight 3 TV Series
The streamer picked up new shows produced with AI technology.
- Another IT governance headache: AI-enabled sanction evasion
Over the next three to five years, both governments and the private sector will need to rapidly adapt identification and mitigation protocols as adversaries move from AI-assisted to AI-enabled sanctions evasion and proliferation financing (PF), a new research paper warns. The report , Algorithms of Evasion: The Rise of AI-Enabled Proliferation Financing, from the Royal United Services Institute ( RUSI ), a UK-based defense and security think tank, defines PF as the use of funds or financial services to acquire, develop or otherwise deal in weapons of mass destruction (WMD). It states, “North Korea and Iran are now developing and deploying AI models to aid with sanctions evasion activities.” Key findings include the fact that AI is now capable of mass producing high-quality fraudulent documents, as well as automating what the report describes as “the administrative minutia of managing extensive shell company networks.” AI powered systems, it states, can also “analyze blockchain patterns in real time to dynamically adjust cryptocurrency mixing strategies, effectively evading detection tools.” In addition, it says, “[tools such as generative AI] which can produce sophisticated fraudulent identification documents, for example, have helped North Korea perpetrate phishing attacks against Western companies.” Dr. Aaron Arnold , senior associate fellow with the Centre for Finance and Security at RUSI, who authored the paper, said in an email that what prompted it was an uptick over the last year in North Korea’s use of AI to facilitate and enhance its cyber operations, in the form of phishing schemes designed to generate revenue for the country’s ballistic missile and nuclear weapons programs. He advised enterprise IT managers who need to protect their organizations from becoming victims of sanction evasion activities that “[it] means largely adapting to a landscape where traditional human-focused security boundaries are being bypassed by automated technologies.” For IT managers, said Arnold, “this might entail incorporating defensive AI, the use of behavior-based analytics, using ‘circuit breakers’ when there is heavy use of API or MCPs, updating personnel training, and hardening identity verification, especially for any remote hiring.” Distinction between AI-assisted and AI-enabled activity is ‘central’ Sanchit Vir Gogia , chief analyst at Greyhound Research, said that the RUSI report matters “because it names the right structural shift. AI is not creating sanctions evasion from thin air, it is compressing and scaling methods that already work.” He pointed out that none of the sanction-evading techniques such as fraudulent documents, synthetic identities, shell companies, hidden beneficial ownership, crypto laundering, and others are new. “What changes is the speed, quality, volume and coordination with which these methods can now be assembled,” he said. According to Gogia, “the distinction between AI-assisted and AI-enabled activity is central. AI-assisted evasion uses AI for discrete tasks: writing a better email, producing a cleaner document, generating a stronger false profile, translating a pitch, summarizing regulations or preparing a plausible job application. AI-enabled evasion is more serious.” A ‘structural asymmetry’ This tactic, he said, “begins to coordinate the system itself. It links identity, documents, ownership structures, payment routes, cloud access, crypto wallets, API calls and timing. The difference is not whether AI helps someone fake a document. The difference is whether AI begins to orchestrate the deception.” That is why the report’s findings should worry enterprise leaders, he noted: “Many organizations still assume the bad actor is mostly human, mostly linear and mostly slow. That assumption is expiring. AI lets adversaries run more attempts, with fewer errors, across more channels, in more languages, with better paperwork and greater patience than most enterprise review processes can absorb. This is not a tale of genius criminals discovering magic. It is the story of ordinary controls meeting industrialized plausibility.” The evidence today, he pointed out, is strongest around tactics such as identity fraud, document fraud, synthetic personas, remote-worker deception, phishing, social engineering, crypto obfuscation and workflow abuse. “Fully autonomous evasion networks sit on the horizon,” he said. “They are serious, but they are not yet the everyday baseline.” This distinction matters, said Gogia: “If enterprises obsess over cinematic autonomous agent scenarios while leaving remote hiring, vendor onboarding, payment approvals, and document review full of holes, they will lose in the most prosaic way imaginable.” The report, he said, also gets the “asymmetry” right. “Offensive actors can learn across the ecosystem,” he said. “They can scrape open information, reuse leaked records, study enforcement patterns, test onboarding forms, inspect public procurement data, watch court filings, probe compliance thresholds and [use the information to] refine their behavior.” Defenders, by contrast, are hemmed in by privacy rules, fragmented data, explainability requirements, jurisdictional boundaries, conservative operating models and siloed technology estates. “Offensive AI learns broadly,” he said. “Defensive AI often learns from fragments. That is the structural asymmetry.” He explained that the regulatory landscape also amplifies the problem, in that regulatory bodies “still speak in separate dialects. [For example] the EU AI Act pushes organizations toward stronger obligations for high-risk AI. NIST-style frameworks push risk management, transparency, and governance.” A trust architecture problem Financial Action Task Force (FATF) expectations push national risk assessment and counter-proliferation controls, he noted, while banking regulators focus on model risk, accountability and operational resilience. “None of these streams is irrelevant. The trouble is that criminals do not organize themselves around regulatory workstreams. They organize around outcomes.” What that means, said Gogia, “is that enterprise cannot wait for a clean global rulebook. It will not arrive in time. CIOs, CISOs, compliance officers and boards need a working governance model now. They need privacy-preserving analytics, controlled data environments, audit trails, legal safeguards and clear model-risk accountability.” He said that enterprise IT managers should treat the situation as a trust architecture problem rather than a narrow sanctions-screening problem. “The uncomfortable truth is that AI is not simply helping bad actors write better phishing emails or forge tidier documents,” he noted. “It is helping them manufacture legitimacy across a chain of enterprise workflows.” Likely outcome an ‘AI arms race’ Report author Arnold also noted that there are signs that cyber criminals have discovered new AI technologies and abilities that legitimate enterprises could adopt for legitimate applications. History, he said, “is replete with [criminals] developing novel solutions to tough problems, [which are] later adopted by law enforcement. Much of our anti-financial crime policy is effectively a response to bad actors exploiting systems or using technology in novel ways to perpetrate crimes. In this scenario, I think an ‘AI arms race’ between enforcement authorities and bad actors is the most likely outcome.” Gogia added, “the baddies are not teaching enterprises how to invent AI. They are teaching enterprises where trust is leaking. That is the lesson worth taking seriously.”
Score: 57🌐 MovesMay 27, 2026https://www.cio.com/article/4177854/another-it-governance-headache-ai-enabled-sanction-evasion.html - Groupon layoffs today: Hundreds of jobs slashed in latest ‘AI-native’ tech company pivot. Stock price rises
Another tech company has announced that it will lay off a significant number of workers in an effort to become “ AI -native.” This time around, it’s Groupon, the legacy discount e-commerce platform that rose to prominence in the early 2010s. Here’s what you need to know about Groupon’s layoffs and its “Project Foundry” AI plans. What’s happened? Last Thursday, the board of Groupon, Inc. (Nasdaq: GRPN) approved a restructuring plan that will see mass layoffs at the company. This information comes from a Form 8-K filing filed with the U.S. Securities and Exchange Commission (SEC) on May 21. In the filing, Groupon revealed it will reduce “up to 400 positions globally.” Those positions include both employees and contractors, and the cuts are expected to happen by the end of Groupon’s Q3 2026. The company is currently in its fiscal Q2 2026, which ends on June 30. Groupon’s fiscal Q3 runs from July 1 to September 30, which means the workforce reductions should occur by October. In a Schedule 14A Proxy Statement filed with the SEC on April 28, Groupon revealed that it had approximately 1,734 employees, which included “full-time, part-time, seasonal and temporary employees.” Groupon said that the figure excluded independent contractors. It is unknown how many contractors are included in the workforce reductions of up to 400 individuals. If the layoffs were to encompass only the company’s employees, they would represent roughly 23% of its employed workforce. Groupon emphasized that this is just the “initial phase” of the restructuring and that more changes are expected. As for why Groupon is cutting jobs—yes, you can once again blame AI. Groupon’s AI-native “Project Foundry” In its 8-K filing, Groupon said that the layoffs were the first part of the company’s strategy to rebuild itself “as an AI-native company,” the goal of which is to “better deliver on our mission, serving both customers and merchants.” As is usually the case with this kind of corporate initiative, the strategy has a name: Project Foundry. As announced in its earnings results on May 7, Groupon describes Project Foundry as a “company-wide initiative to transform our operating model by embedding AI agents into the core of every function.” Additionally, it says the initiative is “intended to enable the Company to operate with the speed required to succeed in an AI-native world.” In other words, Groupon’s Project Foundry is the company’s attempt to use AI agents in place of human workers in everything from marketing to sales. In that regard, despite its name, Project Foundry is no different from what other legacy tech companies are doing as of late: looking for ways to replace humans with AI bots to save costs and, hopefully, increase sales. And speaking of saving costs, Groupon said it expects its layoffs of around 400 workers to eventually save the company between $20 million and $25 million each year. Those cost savings are expected to be around $10 million to $12 million in 2026, with the company planning on reinvesting up to 50% of those savings this year “in marketing, AI infrastructure, and talent density.” In its 8-K filing, Groupon also said it “is currently evaluating additional material cost-reduction and automation actions related to Project Foundry.” What those “automation actions” entail is unknown, but Groupon says that it “expects any such actions would be completed by the end of 2027.” Groupon’s stock rises on AI-native push Investors lately have tended to reward these types of AI-driven pivots, and so it’s little surprise that Groupon’s stock price is rising after the news of AI-driven layoffs. Yesterday, GRPN stock surged more than 9% to close at $20.69—a high for the year. As of the time of this writing, in premarket trading this morning, GRPN shares are up another 1.4%. Based on the company’s closing share price yesterday, Groupon’s stock is now up over 17% in 2026. However, the company’s stock price is still down around 22% over the past 12 months. And looking back even further, things are much worse for the company’s stock price. Groupon was one of the early 2010s internet darlings, and the company went public in 2011. But in the time since, the stock has fallen by more than 96% as Groupon’s popularity waned. Now the company is clearly hoping that an “AI-native” push could see it reclaim some of its lost glory. But whether that happens, as with all things AI, remains to be seen.
- The emerging AI content licensing market puts news publishers in a “double bind,” a new report warns
A new report from the thinktank Open Markets Institute scopes out the current state of AI content licensing for news publishers. “Same Gatekeepers, New Tollbooths: Mapping the AI Content Licensing Market” explores the emerging market for content licensing, arguing that news publishers are currently in a “double bind”: The same big tech companies that are...
- TSMC, IBD Stock Of The Day, Flirts With Buy Point As Chip Giant Will Raise Prices Amid AI Demand
TSM stock rose Wednesday after Taiwan Semiconductor Manufacturing announced plans to raise prices on its leading-edge chips. The post TSMC, IBD Stock Of The Day, Flirts With Buy Point As Chip Giant Will Raise Prices Amid AI Demand appeared first on Investor's Business Daily .
Score: 56🌐 MovesMay 27, 2026https://www.investors.com/research/ibd-stock-of-the-day/tsm-stock-buy-point-tsmc-chip-giant-raise-ai-chip-price/ - CERT-In releases blueprint for defending against AI-assisted cyber threats
CERT-In has released a new blueprint detailing how organisations should defend against AI-assisted cyber threats, including deepfake attacks and AI-generated malware. The post CERT-In releases blueprint for defending against AI-assisted cyber threats appeared first on MEDIANAMA .
- Anthropic Is Likely Generating at Least 35% More Revenue Than OpenAI
Anthropic Is Likely Generating at Least 35% More Revenue Than OpenAI The Information
Score: 56🌐 MovesMay 27, 2026https://www.theinformation.com/newsletters/the-briefing/anthropic-likely-generating-least-35-revenue-openai - China’s Industrial Profits Grow at Fastest Pace in Over Two Years on AI Boom
China’s Industrial Profits Grow at Fastest Pace in Over Two Years on AI Boom Caixin Global
- GPT-5.5 Beats Claude and Gemini in New Long-Horizon Coding Benchmark
DeepSWE is a new benchmark for testing real-world AI coding capabilities.
Score: 55🌐 MovesMay 27, 2026https://analyticsindiamag.com/ai-news/gpt-55-beats-claude-and-gemini-in-new-long-horizon-coding-benchmark - Samsung Unions Approve Pay Deal That Highlights Inequality of A.I. Age
The agreement all but guarantees hefty bonuses for employees in the top-performing chip unit. Other workers say they feel left out.
Score: 55🌐 MovesMay 27, 2026https://www.nytimes.com/2026/05/27/world/asia/samsung-ai-profit-bonus-workers.html - China Wants Its Companies to Embrace AI—Without Firing Workers
As a backlash against AI builds in the U.S. and elsewhere, China is acting to stave off social and economic disruption.
- Xiaomi Slashes AI Model API Prices by 99% to Match DeepSeek
Xiaomi Slashes AI Model API Prices by 99% to Match DeepSeek Caixin Global
Score: 55🤖 ModelsMay 27, 2026https://www.caixinglobal.com/2026-05-28/xiaomi-slashes-ai-model-api-prices-by-99-to-match-deepseek-102448357.html - Google folds Display Ads into AI-first Demand Gen platform
Google is folding Display Ads into its AI-powered Demand Gen platform, marking the end of a long-standing digital advertising model. The Google Display Network (GDN) has been a staple of the open internet for almost twenty years. Marketers previously relied on its predictable framework to target placements, bid on audiences, and A/B test static creative […] The post Google folds Display Ads into AI-first Demand Gen platform appeared first on AI News .
Score: 55🌐 MovesMay 27, 2026https://www.artificialintelligence-news.com/news/google-folds-display-ads-ai-first-demand-gen-platform/