METR

METR is a safety research organization focused on evaluating and understanding the capabilities and risks of advanced AI systems, particularly large language models and frontier AI. The organization conducts empirical research on AI alignment, interpretability, and autonomous AI behavior to help inform governance decisions and corporate safety practices. METR's work is significant for enterprise compliance because it provides technical evidence about AI system vulnerabilities and failure modes that organizations need to understand when implementing responsible AI policies.

1 item

ResearchGlobal2026-08-03

MirrorCode Benchmark Shows AI Can Autonomously Build 16,000-Line Codebases

Epoch AI and METR published MirrorCode, a benchmark measuring how large a software project an AI agent can autonomously reimplement without access to the original source code. Tasks in the benchmark ran for up to 19 days and cost up to $2,600 per attempt. Claude Opus 4.7 successfully completed the benchmark's largest evaluated task, reimplementing a 16,000-line bioinformatics toolkit in 14 hours.

Claude Opus 4.7 Anthropic Epoch AI METR agentic AI benchmarking autonomous agents AI capability evaluation