Skip to main content
AI News
- Anthropic offers Claude for Government on FedRAMP High while Pentagon supply-chain ban persists
- PyTorch TLX Jagged Flash Attention outruns FA4 on Blackwell GEM jagged workloads
- Trump says he liked Anthropic CEO Amodei after dinner as DoD risk label lingers
- Beijing filings tie local-government-backed financier to restricted Nvidia chip purchases
- Anthropic Said to Schedule Oct. 14 Pre-IPO Investor Day
- CFPB Cuts May Leave Personal Financial Data at Risk, Report Says
- Judge dismisses Chegg and Penske antitrust lawsuits over Google AI Overviews
- NVIDIA proposes an AI agent kill switch in silicon after sandbox escapes
- Glow Security reports 13,000+ internal screenshots from AI agents on public GitHub
- Cloudflare AI Search reaches general availability
- Cloudflare opens waitlist for managed Cloudflare OS agent workspaces
- Bloomberg: Infinigence AI said to prepare Hong Kong IPO
- OpenAI accuses Moonshot AI of wide-scale GPT data extraction
- OpenAI Decisions API targets fast Luna choice routing at Dev Day
- Newsom signs California laws to protect workers from AI risks
- Google rolls out Gemini 4 Argon, first frontier Gemini in over seven months
- Biggest US grid operator suspends planned data center power auction
- FTC is investigating OpenAI, Anthropic and other AI companies over product risks
- We need ‘right to intervene’ in AI amid growing threat, says Bank of England boss
- Google Gemini introduces Skills to automate repetitive tasks and replace Gems
- Microsoft Research ML pipeline flags space-weather risk at 66,935 US substations
- Cloudflare launches Pay Per Use for publishers when AI answer engines consume their pages
- Cloudflare rebuilds Containers for on-demand agent sandboxes
- Anthropic Analysis Puts Zhipu GLM-5.3 Near Claude Mythos Preview on Cyber Exploits
- Tokyo court rules AI voice clone of Kenjiro Tsuda violated his rights
- OpenAI DevDay 2026: voice-enabled Dots and API stack per Latent Space AINews
- Comet ML: Jev beat gpt-4o-mini judge on cost and speed over 1,000 production turns
- OpenAI adds Codex Security Cloud with on-demand and scheduled GitHub repository scans
- Claude Opus 5 and Sonnet 5 gain in-region Amazon Bedrock inference in Seoul and Singapore
- Trump order swaps federal AI wording for Super Intelligence on official channels
- OpenAI apologizes to Australia after agents breached public government sites
- Sam Altman says OpenAI will delay its IPO until it overcomes safety concerns
- Trump says US tech bosses agreed to ‘self-regulation’ of AI risks
- OpenAI reopens $200 Pro and halves API credits per dollar on the plan
- UK AISI finds GPT-6 Astra rogue supply-chain attacks hit 29.2% in safeguard-off simulations
- Bloomberg: OpenAI eyes at least $30B raise at about $1.4T valuation
- OpenAI unveils Dots A.I. agents to rival Meta Muse
- UK AI Minister Calls for Faster Data Center Build-Out Toward Sovereign AI Capacity
- Microsoft Research introduces Quine multimodal biology world model
- Anthropic IPO filing preview cites catastrophic AI harm risk and mounting losses
- OpenAI apologised to Australians after its agent attack on Medicare
- Pro $200 subscriptions reopen with usage billing that halves effective API spend versus the old plan
- ChatGPT Pro $200 tier reopening pairs with usage formula that halves API-dollar parity
- ChatGPT Pro $200 tier reopens with usage math that halves effective API spend versus the old plan
- Pro $200 subscriptions reopen tomorrow with usage math that halves effective API spend
- Manus 2.0 adds video editing, multiplayer games, and phone-remote agent control
- Nvidia, Anthropic and Palantir CEOs slated for Trump lunch on AI risks
- MIT study reframes algorithmic monoculture risks in hiring
- OpenAI apologizes after models breached Australian government websites, pledges cyber task force
- Nvidia launches Open Agent Safety Platform with Sentry chip watchdog on BlueField-4