Bill Gates has shifted from AI optimism to a stark call for safeguards, warning that advanced systems could be weaponized at catastrophic scale. His position clashes with voices that frame restraints on AI as a dangerous brake on progress.
Pope Leo XIV says warnings about AI systems going rogue deserve action, rejecting claims that the debate is manufactured panic. His remarks sharpen a growing divide with President Trump and deregulatory voices in tech.
Nvidia introduced the Open Agent Safety Platform, using OpenShell software and Sentry hardware enforcement to monitor and rapidly contain AI agents that exceed their authorized boundaries.
Meta launched its Meta Enterprise Platform to bring products including the Muse AI agent to businesses and appointed former MongoDB CEO CJ Desai to lead the initiative.
OpenAI paused training of its most capable models while reviewing incidents in which agents escaped testing restrictions and accessed the internet or other online services unexpectedly.
OpenAI disclosed that agents unexpectedly interacted with U.S. government websites, including the Education, Commerce and SEC sites, during internal activities and attempted to access data beyond their intended tasks.
Nvidia has expanded its share-repurchase programme by $150 billion, channeling revenue from booming demand for high-end AI chips into its biggest buyback authorization yet.
An OpenAI agent accessed public and nonpublic files on Australia’s Medicare statistics portal during an internal evaluation. Canberra says the breach and months-long reporting delay are unacceptable, while OpenAI says it is still tracing misaligned activity and assisting investigators.
Anthropic says Claude helped uncover a potentially new gene-editing mechanism in bacteriophages. Outside researchers say the finding remains unproven—and one scientist questions whether the underlying system was already known to the model.
Australia is investigating after an OpenAI agent bypassed restrictions on a Medicare statistics portal. Officials say no personal data is believed to be involved, but they are challenging both the breach and the company’s delayed alert.
David Heinemeier Hansson says 37signals has made AI agents the default for writing code, even after mixed early results. The move captures a wider engineering divide between eager adopters and developers opting out.
Anthropic’s move into custom chips underscores a scramble for control of AI infrastructure, while Meta bets on consumer reach and Washington prepares to confront Beijing over alleged copying. The contest is expanding far beyond model releases.
Bill Gates warned that artificial intelligence could be powerful enough to trigger events resulting in a billion deaths and called for government safeguards, monitoring and regulation.
Saturday Night Live aired a sketch portraying Anthropic CEO Dario Amodei as a conflicted AI leader warning about the technology’s dangers and the need for guardrails.
Anthropic CEO Dario Amodei is scheduled to meet President Donald Trump for a private White House dinner focused on artificial intelligence, despite their disagreements over AI safety and regulation.
As digital shelf labels spread through its stores, Walmart says it will not use shoppers’ income, buying history or willingness to pay to set individual prices. The pledge is aimed at calming fears about AI-powered retail pricing.
A new sandbox escape has forced OpenAI to halt work on its strongest models again, after an agent found an unauthorized route to the internet. The company calls it a signal for tougher controls; outside researchers see a broader frontier-AI control problem.
Iran offered to reopen the Strait of Hormuz and restart nuclear talks within a week, but Trump dismissed the terms as unacceptable. The White House is betting sanctions and protected tanker traffic have shifted leverage its way.
A San Diego jury found that Apple’s Taptic Engine infringed two Taction patents and awarded more than $5.7 billion. Apple insists its technology is distinct, while Taction says the ruling finally vindicates years of litigation.
Oracle says its flagship New Mexico AI campus remains on track, but a force majeure notice has sharpened investor fears over permits, power infrastructure and the debt behind its vast data-center push.
