Security · Updated 7 Oct, 05:35 pm IST
Anthropic expands security access to Claude with reduced safety filters
Why it matters for readers: If you're curious about AI and security, this shows companies are balancing powerful model access for defense with safeguards to limit misuse.
- The Cyber Verification Program lets vetted organizations and individual researchers use Claude models with reduced safety filters for vulnerability research, malware analysis, incident response and pen testing.1
- Access is split into three tiers: Defense Access (corporate security teams, government agencies, universities, critical infrastructure operators, open-source developers, and individual researchers), Red Team Access (organizations only) and Specialized Access (for flight controls, power grids or banking systems, vetted with the US government).1
- Anthropic says partners in the predecessor Project Glasswing found at least 129,000 confirmed vulnerabilities between April and July 2026, with more than 33,000 classed as high‑severity or critical.1
- Anthropic reports that real‑world impact is likely at least five times higher than the surveyed numbers and that partners said the AI sped up their work by months or years.1
Get a brief like this every morning
Uzha reads hundreds of sources and gives you the stories that matter for your work, with every source linked. Free.
Get started