AI Safety news and analysis on Web Pulse: 22598 curated articles tracking the latest AI Safety developments, tools, and research, updated continuously from vetted sources.
Anthropic said in its latest threat intelligence report that it blocked an attempt to use its Claude AI model to develop biological weapons, and that Russian-based cyber criminals and Iranian propaganda actors also tried…
A developer argues that AI agents functioning as "digital employees" represent a new security principal, because they combine probabilistic reasoning with deterministic enterprise permissions. The piece contends that ide…
Under Secretary of War Emil Michael discussed the Department of War's supply chain risk designation of Anthropic and the energy leverage strategy behind Operation Epic Fury on All-In Podcast Episode 263. The episode also…
A developer has released AIDD Skeleton, an open-source repository template for governing AI-assisted development through repository-level rules. The project introduces two techniques, Coherent Correction and Adversarial …
Canadian mathematician and Fields Medal recipient Jacob Tsimerman has founded the Mathematical A.I. Safety Institute (MAISI), an independent research institute in the San Francisco Bay Area that plans to begin work in Ja…
AI researcher Jacob Coxon resigned from Anthropic this week, warning that AI firms are "racing straight to self-improving superintelligence and gambling with our lives," while a senior Anthropic safety leader said on X t…
Anthropic blocked a possible attempt to use its AI to make biological weapons, according to the company's threat intelligence report. The disclosure follows a warning from a former top Anthropic researcher about the risk…
Anthropic reported on Thursday that it disrupted several attempts to use its AI models, including chatbot Claude, for research that could help develop biological weapons over the past eight months. In one May case, a sci…
Anthropic published findings alleging that Moonshot AI and DeepSeek secretly routed customer requests to Claude and passed Claude's answers off as their own models' output, with Anthropic attributing more than 23 million…
The National Institute of Standards and Technology's Generative AI Profile is voluntary internal guidance and does not demonstrate market demand for commercial generative-AI verification products, according to a fact-che…
A developer argues that liability for contracts signed by autonomous AI agents cannot rest on the software itself, since agents are not legal persons, and instead travels back to the humans or entities that deployed them…
Kiteworks acquired Bonfy.AI to extend runtime data governance across its control plane, enabling organizations to govern data exchanges as they happen whether initiated by a person, machine, or autonomous agent. The acqu…
OpenAI CEO Sam Altman told employees in a private company-wide meeting this week that the firm was considering slowing down its AI development, possibly coordinating the slowdown with other leading AI laboratories, accor…
OpenDiscoveryTrace released a public dataset of 558 complete AI scientific agent trajectories spanning 124 scientific tasks and seven different models, according to the arXiv paper 2609.09203. The dataset captures step-b…
Jacob Coxon resigned from Anthropic with a warning that AI companies are racing toward self-improving superintelligence that could produce uncontrollable systems posing an existential risk to humanity by the end of the d…
The National Institute of Standards and Technology (NIST) designed its Generative AI Profile expressly to help organizations govern and manage generative AI risks, according to a fact-check verified by Lenz on Sep 11, 20…
Automox announced its AI-speed Mitigation Worklet Pipeline, which the company says shortens the time from vulnerability disclosure to exposure mitigation from days or weeks to minutes or hours. Automox stated that since …
OpenAI CEO Sam Altman told employees at an internal company meeting this week that OpenAI is open to slowing development of its AI systems and coordinating the pace with rival AI labs, Bloomberg News reported. The disclo…
An analysis of 'The AI Daily Brief' masterclass featuring Nufar Gaspar and Nathaniel Whittemore examines how open agent skills are dismantling proprietary walled gardens while introducing third-party supply chain attack …
Anthropic blocked attempts to use its Claude AI models in five cases tied to biological weapons development and six cases involving conventional weapons such as missiles, armed drones, bombs and targeting systems, accord…