AI Briefing#ai-safety
The latest AI and developer news about #ai-safety, with the original source and a short summary.
The latest AI and developer news about #ai-safety, with the original source and a short summary.
OpenAI will participate in the Paris AI Action Summit to discuss ways to drive economic growth through AI and will unveil an economic blueprint for Europe.
OpenAI has released a system card analyzing the safety evaluations and risk factors of the o3-mini model.
OpenAI has released a safety report for 'Operator,' an agent that combines GPT-4o's vision capabilities with reinforcement learning to directly operate a computer.
Anthropic has obtained ISO/IEC 42001 certification, the international standard for AI governance, demonstrating trust in its AI safety practices.
This analyzes the technical definition of AI agents and the safety and ethical risks that come with their increasing autonomy.
Anthropic has developed a multi-pronged response to prevent AI misuse during the 2024 global election season, including strengthened policies, model testing, and provision of accurate information.
Anthropic has proposed that the federal government expand its investment in NIST to accurately measure the capabilities and risks of AI systems.
Anthropic argues that a sophisticated third-party testing framework is needed to ensure the safety of frontier AI systems.
Anthropic identifies and mitigates election-related AI risks through expert-based policy vulnerability testing and automated evaluations.
OpenAI has released a system card detailing the technical architecture of its video generation model Sora and the measures taken to ensure its safety.
OpenAI has released a system card detailing the o1 model's reasoning capabilities and safety evaluation results.
OpenAI announced ways to advance red teaming techniques that combine humans and AI to detect risks in AI models.
As AI advances rapidly, targeted regulation that does not stifle innovation is needed to address cyber and CBRN risks.
Anthropic has updated its **Responsible Scaling Policy (RSP)** to manage catastrophic risks from AI.
OpenAI has disrupted more than 20 deceptive operations since earlier this year and has released its latest threat intelligence report.
OpenAI has introduced a new GPT-4o-based multimodal moderation model into the Moderation API.
OpenAI operates an independent board-level Safety and Security Committee to oversee the safety of model development and deployment.
OpenAI has released the full list of contributors, including researchers and safety teams, who participated in developing the o1 model.
OpenAI has released the list of external red teamers and collaborators who participated in the safety verification of the o1 model.
Anthropic has updated its Usage Policy to strengthen election integrity, high-risk case handling, and privacy protection.
Anthropic shared the achievements and future advancement directions gained from implementing the Responsible Scaling Policy (RSP).
Anthropic is launching a new bug bounty program to find 'universal jailbreak' vulnerabilities in its next-generation AI safety systems.
Jay Kreps, co-founder and CEO of Confluent, has joined Anthropic's board as a member.
Anthropic is expanding the availability of its **Claude** models to the U.S. intelligence community and government agencies through AWS.