X-Risk Daily logo

X-Risk Daily

Archives
Log in
Subscribe
September 2, 2026

OpenAI says new model Astra crosses 'critical' cybersecurity risk threshold

Also: USPS ballot-screening system prompts whistleblower and Democratic accusations of a 'power grab'‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ 

X-Risk Daily

Wednesday 02 September 2026

Transformative AI
OpenAI says new model Astra crosses 'critical' cybersecurity risk threshold
OpenAI confirmed on 1 September 2026 that its upcoming model Astra is the first system it has built to meet the "Critical" cybersecurity capability threshold under its Preparedness Framework, the internal system the company uses to classify and respond to dangerous AI capabilities.

Also in this issue

Fanatical & Malevolent Actors
USPS ballot-screening system prompts whistleblower and Democratic accusations of a 'power grab'
Geopolitics & Conflict
Iran strikes US bases after wedding party deaths blamed on American attack
Transformative AI
Anthropic builds customer-controlled data system to detect AI misuse without retaining logs itself
Geopolitics & Conflict
US strikes Iran's southern coast, denies retaliation hit Jordan bases
Research
Researchers link poor-quality RL training data to AI reward hacking
Research
Frontier AI models fail to in-context learn obscure board game that humans pick up quickly

… and 40 more in the full briefing

Read today's briefing →

Prefer a weekly digest? Click ‘manage your subscription’ below.

Got thoughts on today’s briefing? Just reply to this email — I’ll read and reply!

Generated automatically from dozens of trusted sources including Transformer, Sentinel, Epoch AI, LessWrong, and EA Forum.

Don't miss what's next. Subscribe to X-Risk Daily:
← Newer UK peers push for legal power to shut down runaway AI systems Older → Major AI firms warn of imminent AI-enabled cyberattack wave; White House cuts biosecurity team as AI bio-risk grows
Powered by Buttondown, the easiest way to start and grow your newsletter.