Epoch AI maintains a registry of 85 AI benchmarks spanning mathematics, software engineering, agentic workflows, and games, and has flagged a subset as flawed. The registry includes benchmarks such as…
OpenAI's GPT-6 Astra solved between 97.6% and 98% of the 43 problems in Tier 4 of Epoch AI's FrontierMath benchmark, earning the first-ever "Major Advance" classification from Epoch for an AI-generate…
Epoch AI launched Benchmark Reviews on September 17th and classified nine of its first 15 reviewed benchmark versions as Flawed, four as Verified, and two as Not Enough Info, flagging defects in tests…
Anthropic said on September 17 that its Claude model now "leads" 26% of the AI research and development work inside the company, up from 1% in March on a scale developed by the nonprofit Epoch AI, and…
Anthropic published three measurements on September 17, 2026 to show how fast AI is being built inside frontier labs, reporting that about 30,000 agents run at any one time on its most-used internal p…
OpenAI's GPT-6 Astra topped the Epoch Capabilities Index with a composite score of 166, outranking all 249 models tracked by Epoch AI and surpassing GPT-5.5 Pro at 162 and Anthropic's Claude Fable 5.1…
Epoch AI's AI Data Centers database now tracks 86 large AI facilities worldwide, covering an estimated 46% of all deployed AI compute as of June 2026, up from roughly 22% in early 2024. The tracked si…
OpenAI president Greg Brockman declared the "AGI era" had begun at the launch of OpenAI's GPT-6 Astra, a claim Nvidia CEO Jensen Huang endorsed on X, but AI researchers including New York University p…
Verizon's 2026 Data Breach Investigations Report, drawing on over a billion vulnerability records, found only 26% of confirmed actively-exploited vulnerabilities on CISA's Known Exploited Vulnerabilit…
A study by Epoch AI finds that from 2019 to 2025, data improvements contributed 12.0x compute efficiency gains compared to 3.7x from model improvements, meaning 3.24x more gains came from data at a 1e…
OpenAI reported that its researchers now use the equivalent of 3.1 agent-workdays for every human workday, with the median researcher spending over $600 per day on inference and the 90th-percentile us…
OpenAI reported that by mid-August 2026, its research organization was using 3.1 agent-workdays of runtime for every human workday, with coding agents expanding beyond research and infrastructure code…
OpenAI reported that coding agents now perform 3.1 agent-workdays for every human workday in its research organization as of mid-August, marking a shift from earlier in the year. The company disclosed…
OpenAI announced it has reached its goal of building an automated 'research intern' capable of independently carrying out well-defined research tasks by September 2026, a milestone the company first l…
OpenAI's GPT-6 Astra has surpassed average human efficiency on the ARC-AGI-3 benchmark for the first time, prompting ARC Prize chief François Chollet to move up his AGI forecast, though he stops short…
Epoch AI launched FrontierMath Erdős, a benchmark of 68 unsolved problems selected by mathematician Thomas Bloom from his erdosproblems.com catalog, to rigorously track AI math capabilities. The bench…
OpenAI released GPT-6 Astra, a computer-use model with a 1,050,000-token context window, available only to Trusted Access and Daybreak organizations due to its designation as the first model to reach …
Arena AI's September 2026 evaluation data shows the Elo rating gap between closed frontier AI models and open-weight competitors has widened to 29 points, up from zero in January 2025, with Anthropic'…
OpenAI's Astra model scored 169 on Epoch AI's Epoch Capabilities Index, the highest composite score ever recorded, surpassing GPT-5's 150 and Claude 3.5 Sonnet's 130. Astra achieved perfect scores on …
Epoch AI published a case study on Anthropic to examine whether financing could bottleneck AI compute. The analysis suggests that while capital availability is a growing constraint, it is not yet the …