Anthropic AI capabilities doubled over past year, study finds
Nexus Europa Newsroom
Posted Nov 9, 2026, 11:30 PM UTC · In the digest of 2026-11-09-2330
An independent study by METR found Anthropic's AI models achieved two years of progress in just 12 months, raising safety concerns.
An independent evaluation by Model Evaluation and Threat Research, a non-profit AI safety organization, has revealed a significant acceleration in the development of Anthropic's artificial intelligence models. According to the research, the technology company achieved roughly two years' worth of performance metric advancements within a single 12-month period.
Researchers involved in the assessment warned that such rapid capability growth raises concerns regarding the potential emergence of complex AI systems that could prove difficult to control or manage safely in the future.
Sources: Cumhuriyet