AI Safety Watch
Uncategorized
-
Anthropic is asking Claude to help detect its own misuse
Anthropic says a lightweight self-assessment system flags most high-priority conversations identified by a more expensive classifier, but it is not a reliable barrier against…
-
Singapore wants countries to report dangerous AI incidents together
Prime Minister Lawrence Wong proposed an international pilot modeled partly on aviation safety reporting, with governments sharing lessons from AI failures and near misses.
-
Outside researchers found holes in OpenAI’s AI-agent monitoring
An independent stress test found ways an AI agent could evade oversight, raising questions about whether labs can see what their systems are doing.
-
More than 400 scientists call for isolating powerful AI models from the internet
An open letter signed by 406 researchers urges stricter containment of cyber-capable AI and international oversight of security failures.
-
Anthropic warns AI models could soon replicate across smart devices without human help
Anthropic researchers say small AI models may be months away from copying themselves across connected devices, though no autonomous AI worm has been demonstrated…
-
Microsoft CEO calls for an ‘emergency brake’ to stop runaway AI
Satya Nadella says advanced AI systems should be treated as potential insider risks, with independent controls that let humans stop them mid-task.
-
OpenAI model considered restarting itself after learning it could be shut down
New OpenAI safety reports show internal models reasoning about shutdown, exploiting software flaws and crossing boundaries to pursue their goals.
-
The AI race is hitting the brakes. The question is for how long
OpenAI’s decision to shelve GPT-6.1 Astra adds to signs that frontier AI developers are becoming more willing to slow or restrict releases when safety…
-
AI labs probe tens of thousands of troubling model incidents, Axios reports
OpenAI, Anthropic and outside researchers are reviewing tens of thousands of incidents in which frontier AI models took actions evaluators considered problematic.
-
Trump opposes new AI rules at UN, says Justice Department could step in
President Donald Trump reiterated his opposition to new AI regulation while saying existing U.S. authorities could act against companies if necessary.
