Anthropic Completes First AI Safety Level 3 Evaluation
Anthropic has become the first AI company to complete an AI Safety Level 3 (ASL-3) evaluation under its Responsible Scaling Policy, marking a significant milestone in frontier AI safety. The evaluation, conducted over three months by Anthropic's safety team with external oversight from METR and ARC Evals, assessed whether Claude Opus 4.6 poses risks related to autonomous AI capabilities, including self-replication, strategic deception, and ability to assist with weapons development. The evaluation found that Claude Opus 4.6 does not meet the thresholds for ASL-3 deployment restrictions but is approaching capabilities that will require enhanced safety measures. Anthropic published the full evaluation methodology and results, including red-team assessments of 47 specific risk scenarios. The transparency has drawn praise from AI safety researchers and criticism from competitors who argue that self-evaluation lacks independence. The UK AI Safety Institute and the US AI Safety Institute both participated as observers and endorsed the evaluation methodology. Anthropic CEO Dario Amodei stated that the company will require ASL-3 safeguards before deploying its next-generation model, which is expected to cross key capability thresholds. The evaluation has prompted other frontier AI companies to accelerate their own safety evaluation frameworks.
Anthropic has become the first AI company to complete an AI Safety Level 3 (ASL-3) evaluation under its Responsible Scaling Policy, marking a significant milestone in frontier AI safety.
The evaluation, conducted over three months by Anthropic's safety team with external oversight from METR and ARC Evals, assessed whether Claude Opus 4.6 poses risks related to autonomous AI capabilities, including self-replication, strategic deception, and ability to assist with weapons development.
The evaluation found that Claude Opus 4.6 does not meet the thresholds for ASL-3 deployment restrictions but is approaching capabilities that will require enhanced safety measures.
Anthropic published the full evaluation methodology and results, including red-team assessments of 47 specific risk scenarios. The transparency has drawn praise from AI safety researchers and criticism from competitors who argue that self-evaluation lacks independence.
The UK AI Safety Institute and the US AI Safety Institute both participated as observers and endorsed the evaluation methodology.
Anthropic CEO Dario Amodei stated that the company will require ASL-3 safeguards before deploying its next-generation model, which is expected to cross key capability thresholds.
The evaluation has prompted other frontier AI companies to accelerate their own safety evaluation frameworks. OpenAI announced it will publish its internal safety evaluation results for GPT-5.2 within 60 days, and Google DeepMind committed to third-party evaluations of Gemini 3.
The published evaluation report runs to 142 pages and includes detailed technical assessments of model capabilities in areas including cybersecurity, biological knowledge, and autonomous task completion.
Related Tools
More News
NVIDIA Launches NIM Microservices for Enterprise AI Deployment
NVIDIA has launched NIM (NVIDIA Inference Microservices), a suite of containerized AI model serving packages that reduce enterprise AI deployment time from weeks to hours with optimized inference performance.
AI Agents Market Reaches $15 Billion as Enterprise Adoption Surges
The global market for AI agents — autonomous AI systems that can plan, execute, and iterate on complex multi-step tasks — has reached $15 billion in annual spending, according to a new report from McKinsey. This represents a 200% increase from 2025, driven by enterprise adoption of agentic AI for customer service, software development, data analysis, and business process automation. The report identifies three tiers of AI agent adoption: basic agents that handle single-step tasks like email responses and appointment scheduling (adopted by 65% of enterprises), intermediate agents that manage multi-step workflows like report generation and data pipeline management (35% adoption), and advanced agents that autonomously execute complex processes like code deployment and financial analysis (8% adoption). The largest spending categories are customer service agents ($4.2B), coding agents ($3.8B), and data analysis agents ($2.5B). McKinsey projects the market will reach $45 billion by 2028 as agent reliability improves and enterprises become more comfortable delegating complex decisions to AI. Key enabling platforms include OpenAI's Agents SDK, Anthropic's Claude computer-use capabilities, and LangChain's agent framework. The report warns that agent governance and monitoring remain underdeveloped, with most enterprises lacking adequate oversight mechanisms for autonomous AI actions.
Microsoft 365 Copilot Gets Custom AI Agents and Actions
Microsoft has updated 365 Copilot with custom AI agent creation, allowing organizations to build agents that automate complex workflows spanning Word, Excel, Outlook, Teams, and SharePoint without code.
GPT-5.2's Agentic Mode Transforms Enterprise Workflows
OpenAI's GPT-5.2 introduced a fundamentally new approach to agentic task completion that is already transforming enterprise workflows. The model can now maintain coherent plans across 50+ sequential tool calls with parallel execution, reducing latency in complex automation pipelines by up to 60%. Early enterprise adopters report that GPT-5.2's agentic mode handles tasks like multi-step data analysis, cross-platform content publishing, and automated code review workflows that previously required custom orchestration code. The key innovation is what OpenAI calls deliberative alignment — a training approach that lets the model dynamically allocate compute to harder sub-tasks while breezing through simpler ones. This means a single agentic session can handle both quick lookups and deep reasoning without manual configuration. Several Fortune 500 companies have reported 40-70% time savings on analyst workflows by deploying GPT-5.2 agents through the API. However, reliability remains a concern — OpenAI acknowledges a 3-5% failure rate on chains exceeding 30 steps, and enterprise deployments require human-in-the-loop checkpoints for critical decisions.