Anthropic Discloses Internal Metrics: Claude Now Drives 26% of Its R&D

On September 18, Anthropic published three internal metrics on its official blog, giving the public a first systematic look at how "AI is being used to build next-generation AI" — a first among frontier labs. Internal snapshots show Claude now drives about 26% of Anthropic's R&D, with roughly 30,000 agents running concurrently.

Confirmed

• Anthropic released three metrics: share of AI R&D done by AI, supervisability of AI agents, and compute allocation; it stated any frontier developer could publish the same metrics and that they can be verified by third parties
• The work comes from Anthropic's economic research series on measuring internal AI development speed, and uses Epoch's six-level automation scale: Claude "leading" corresponds to AL4 — end-to-end completion of most tasks under high-level prompts, still under human supervision; the share rose from about 1% six months ago to about 26% as of August, with over 90% of R&D work at "AI collaboration" (AL3) level or above; no subtask has reached full L5 autonomy
• Posts differ slightly on the starting point: under 1% in February (@Polymarket, @DrSingularity), about 1% in March (@markk), or zero seven months ago (@AxomaticallyExtinct)
• About 30,000 agents run concurrently on Anthropic's most-used internal platform for research and engineering
• Per @gaganghotra, only 6% of AI R&D compute goes to safety
• Per @chrismattmann citing the official blog, since May about 80% of code written for Claude products was actually generated by Claude itself
• Per @erikbryn, a Stanford-related panel is proposed as a third-party auditor for external validation
• The Washington Post, AP, and Business Standard covered the disclosure, all citing the 26% figure; @TansuYegen argued it signals the trend better than another benchmark score; Polymarket opened a prediction market on it

Unconfirmed

• User shawnup questioned the wording motive behind the "labs in democratic countries" qualifier, a personal view without Anthropic response
• Commenter AryHHAry noted the work is narrower than reproducible science, covering only speed, supervision, and compute

Why it matters

• This is the first time a frontier lab has quantified AI participation in its own R&D, providing a trackable public baseline for assessing AI self-improvement, though reproducibility is questioned
• The jump from about 1% to 26%, combined with 30,000 concurrent agents and about 80% of product code AI-generated since May, suggests AI development speed may decouple from the human-research paradigm — a key datapoint for assessing intelligence-explosion risk; @markk notes the feedback loop has formed even without full autonomy
• The 6% safety compute share has also fueled debate over whether safety investment matches acceleration

Timeline

• February (or March): Claude-led internal AI R&D share under 1% (or about 1%)
• From May: about 80% of code for Claude products generated by Claude
• August: share reaches about 26%; over 90% of work at AI-collaboration level
• September 18: Anthropic publishes the three metrics and internal snapshots

2026-09-18 ~ 2026-09-19 · 41 related posts

Full story(4 episodes)→

• Episode 1: Anthropic: Claude Now Writes 80% of Its Engineers' Code(2026-09-13, 2 posts)
• Episode 2: Anthropic Discloses Internal Metrics: Claude Now Drives 26% of Its R&D(2026-09-18, 41 posts)
• Episode 3: Anthropic Unveils Three Public Metrics to Track AI Progress(2026-09-18, 2 posts)
• Episode 4: Anthropic Says Claude Now Drives 26% of Its Internal AI R&D(2026-09-20, 3 posts)

• [source] Anthropic Publishes Three Internal Metrics to Track AI R&D Automation and Oversight — AnthropicAI · 2026-09-18
• Anthropic publishes measurements of AI development pace inside frontier labs — ResultBackground2450 · 2026-09-18
• [source] Anthropic reveals Claude now leads 26% of its own R&D, up from near zero 6 months ago — Outside-Iron-8242 · 2026-09-18
• Users Question Anthropic's Transparency Report: Why the 'Democratic Countries' Qualifier? — shawnup · 2026-09-18
• Anthropic Publishes First Transparency Metrics: Claude Leads ~26% of Its Internal AI R&D — TheZachMueller · 2026-09-18
• Anthropic: Claude Now Handles 26% of Internal AI R&D, With 30,000 Agents Working at Any Time — Dr_Singularity · 2026-09-18
• Claude now leads 26% of Anthropic's model R&D, 30,000 agents at work — ChrisGPT · 2026-09-18
• Anthropic: Claude now leads 26% of its AI R&D, up from under 1% in February — Polymarket · 2026-09-18
• Anthropic discloses 6% of AI R&D compute goes to safety, publishes three transparency metrics — gaganghotra_ · 2026-09-18
• Anthropic's three AI-progress metrics get a sober critique: disclosure, not reproducible science — AryHHAry · 2026-09-18
• Anthropic says Claude now leads 26% of its R&D, with ~30,000 internal agents — shiringhaffary · 2026-09-18
• Anthropic: 80% of code is written by Claude, and CI jobs exploded 25x in six months — 新智元 · 2026-09-18
• Anthropic says Claude is taking over work of building its own successor — Evening_Bicycle3113 · 2026-09-18
• Anthropic says 80% of Claude's own code is now written by Claude — chrismattmann · 2026-09-18
• Claude Now Leads Over a Quarter of Anthropic's Own AI R&D Work — TansuYegen · 2026-09-18
• [source] Claude now leads 26% of work building its own successor, up from 0% seven months ago — AxomaticallyExtinct · 2026-09-19
• Anthropic: Claude is helping build its own next-generation successor — ThereWas · 2026-09-19
• Claude now leads 26% of Anthropic's AI R&D while 30,000 agents run concurrently — krishnan · 2026-09-19

23 near-duplicate retellings: NathanpmYoung · Polymarket · rohanpaul_ai · shiringhaffary · 新智元 · 机器之心 · Hesamation · APPSO · AI寒武纪 · erikbryn · repligate · mark_k · 002Chris · 002Chris · ZeroStateReflex · MrsSynchronie · rohanpaul_ai · vyxex · anshulkundaje · VraserX · kuanhoong · emmanuelvivier · emmanuelvivier