OpenAI and Anthropic Warn Time Is Running Out for AI Threats
OpenAI and Anthropic call for urgent action by governments and firms against AI cyber risks.
Posts reference an Axios report on an open letter from OpenAI and Anthropic. The letter states that governments and organizations must prepare now for AI-driven cyber threats, with limited time remaining. Sam Altman posted that the challenge requires intense coordinated global response and invited collaboration. Jeffrey Ladish replied that open-weight models currently favor attackers, as vulnerabilities are found faster than they can be patched. Other accounts shared Altman's message and noted the call for collective cyber defense efforts.
JUST IN: OpenAI, Anthropic warn time is running out to prepare for AI threats https://www.axios.com/2026/08/27/openai-anthropic-issue-dire-cyber-threat-warning
Combined views
100.9K
Reactions
6 postsDF 
Danielle Fong 🔆@DanielleFongsam altman says cybersecurity apocalypse time
- likes: 66
- replies: 8
- bookmarks: 5
- reposts: 3
“P “paula”@paularamblesSam if ever there were a time to hit Shift
- likes: 1.1K
- replies: 17
- bookmarks: 48
- reposts: 10
JL Jeffrey Ladish@JeffLadish"We have a limited window to strengthen cyber defenses" = people are about to get wrecked by cyber criminals using extremely capable open weight models to compromise an unprecedented number of people's computers, companies and individuals alike. Get ready!
- likes: 34
- replies: 4
- bookmarks: 7
- reposts: 5
JL Jeffrey Ladish@JeffLadishYes you can use open weight models to defend against attacks too. But in the short term the balance is tipped towards offense. People are finding vulnerabilities in critical software faster than maintainers can patch. Patching is harder to automate than exploitation
- likes: 0
- replies: 0
- bookmarks: 1
- reposts: 0
RB RB