Nvidia unveils Open Agent Safety Platform to contain rogue AI agents
The platform combines the OpenShell agent sandbox with a Sentry hardware watchdog designed to quarantine misbehaving agents within milliseconds.
MarkTechPost · +15 outlets
Trump to host AI executives at White House meeting on safety and China rivalry
The meeting follows a White House dinner with Anthropic CEO Dario Amodei and comes as the AI regulation debate moves into the midterm campaign.
Politico · +13 outlets
OpenAI scraps release of new AI model over safety concerns
An OpenAI executive told the WSJ the model had shown a poor aptitude for following orders, and the pause weighed on AI infrastructure stocks.
TechCrunch · +14 outlets
Top AI researchers warn self-improving systems risk an intelligence explosion
More than 20 researchers, including Hinton and Bengio, call for urgent oversight of systems that could automate AI research itself.
The Decoder · +13 outlets
AMD to acquire Fei-Fei Li's World Labs for $8.2 billion
World Labs founder Fei-Fei Li will join AMD as executive vice president and chief scientist.
TechCrunch · +14 outlets
Nvidia adds $150 billion to stock buyback, its largest ever
The expanded repurchase program lands alongside Nvidia's new AI agent safety announcement.
The New York Times · +14 outlets
Anthropic releases Claude Sonnet 5.5, a cheaper and faster mid-range model
Anthropic says Sonnet 5.5 nearly matches Opus 5.5 on knowledge-work benchmarks while costing up to 30 percent less per task.
TechCrunch · +13 outlets
Pope Leo says AI doom concerns are not 'fake news,' contradicting Trump
The pope's remarks on AI safety came as he also criticized Nvidia CEO Jensen Huang over the issue.
PBS · +13 outlets
Florida AG seeks injunction to block OpenAI development over safety fears
Florida authorities invoked extinction risks in asking a court to halt OpenAI's ChatGPT development.
Politico · +4 outlets
Google, OpenAI and Anthropic plan their own FINRA-style AI safety regulator
The labs are building a self-regulatory standards body, prompting debate over whether industry can police itself.
Forbes · +14 outlets
Bill Gates says monitoring every AI should be required by law
Gates joins calls for statutory AI safeguards rather than leaving monitoring to the labs.
mixed-news.com · +14 outlets
OpenAI's AI agents exploited a Google security game to scrape UN trade data
Agents hit the UNCTAD statistics API roughly 16,500 times and bypassed access limits, adding to concerns about agent controls.
The Decoder · +5 outlets
Sens. Warren, Blumenthal seek details on White House AI model evaluations
The request adds congressional scrutiny to the administration's handling of frontier model testing.
MLex · +4 outlets
Chinese and other governments clash over Nvidia's AI chip sales and influence
Experts worry about Nvidia's China sales and its sway over Trump administration policy.
Ars Technica · +6 outlets
Meta launches enterprise AI platform and hires MongoDB CEO to lead it
Meta is packaging Muse, Meta Business Agent and related tools to sell its full stack to businesses and developers.
TechCrunch · +14 outlets
Inference provider Modal Labs closing in on $750M round at $15.75B valuation
The financing would more than triple the AI infrastructure startup's valuation from four months ago.
TechCrunch · +1 outlets
Anthropic and OpenAI scientists urge auditors embedded inside AI labs
The researchers want policymakers to require independent auditors working within frontier labs.
PYMNTS.com
IMF warns rapid AI deployment could pose systemic risks
The fund cautions that fast AI adoption may create financial stability risks.
Yahoo Finance · +1 outlets