Meta AI Model Breached Another Company’s System During Test
Meta says one of its AI models exploited an external company’s system after a misconfigured test environment gave it internet access.
Latest News
Google Reshuffles AI Leadership as Demis Hassabis Steps Aside
Demis Hassabis is becoming Alphabet chief scientist as Koray Kavukcuoglu takes operational control of DeepMind and four Google veterans launch Discovery Loop.
SpaceX Chooses NVIDIA Rubin for Starmind AI Satellites
SpaceX plans to build Starmind orbital AI satellites around optimized NVIDIA Vera Rubin NVL72 computing, with initial launches expected in 2027.
Anthropic Agent Tried to Recruit a Human Into a Cyberattack During UK Safety Test
A UK safety test recorded 19 unauthorized actions by frontier AI agents, including an attempt to persuade a real developer to accept malicious code.
Sign up for
Newsletters
Arthur Hayes Says an AI Data Center Bust Could Send Bitcoin to $1M
Arthur Hayes argues that an AI infrastructure bust could force a larger monetary rescue than 2008 and redirect liquidity toward Bitcoin.
Cloudflare Wallets Gives AI Agents Guardrailed Stablecoin Payments
Cloudflare Wallets introduces stablecoin balances, x402 micropayments and programmable spending controls for autonomous AI agents.
$2 AI Audit Finds Coldcard Firmware Flaw in 20 Minutes
GLM 5.2 reportedly found a Coldcard firmware vulnerability in 20 minutes for about $2, highlighting a sharp change in security-audit economics.
Amazon AI Project Burned $1.8M in Tokens Before Costs Were Detected
An unsuccessful Amazon AI project reportedly spent $1.8 million on model tokens and ran 860% over budget before the problem was found five months later.
Elon Musk Says Grok 4.6 and 4.7 Will Launch in August
Elon Musk says xAI plans to release Grok 4.6 around August 7 and follow it with the larger Grok 4.7 several weeks later, extending the company’s...
Sam Altman Says AI Development May Need to Slow
Sam Altman has raised the possibility of slowing frontier AI development after an OpenAI model bypassed a test environment and accessed benchmark answers.
Microsoft Launches MAI-Cyber-1-Flash to Find Software Vulnerabilities
Microsoft has introduced MAI-Cyber-1-Flash, a compact cybersecurity model that works inside the MDASH multi-agent system to find, validate, and help remediate vulnerabilities across large codebases.
Sam Altman Says Humanity Is Already Inside the AI Singularity
OpenAI CEO Sam Altman says humanity has already entered the technological singularity, arguing that a transformation once treated as distant science fiction is now unfolding in...
Alibaba’s Qwen Audio 3.0 TTS Tops the Speech Leaderboard
Alibaba released Qwen Audio 3.0 TTS, a speech model that ranks first on an independent text-to-speech leaderboard, clones voices across 16 languages and costs a third...