Anthropic has released a set of internal measurements aimed at making the pace of AI development more transparent to the public, covering three areas: how much AI R&D is performed by AI itself, how effectively AI agents are overseen, and how compute is allocated. The company shared a snapshot of these metrics from its own operations and says any frontier developer could publish equivalent measures, with third-party verification possible. Anthropic argues the gap between what frontier labs and the public know should be minimized as society weighs how to pace frontier development.

AI summary of the linked source — not editorial content.

Story signal · how it spread on X · via Elfa

as of
thread quiet
Spread
i
Distinct X and Telegram accounts Elfa counted behind this story’s reports, summed over the thread. Not posts, not impressions.
40 accounts

first post
Leviathan +1.6 h later

ANTHROPIC reaction
i
Price change from the first post to 24 hours later (Hyperliquid). BTC over the same window is shown so a market-wide move is not mistaken for a reaction to this story. Not advice.
+1.3% 24h after post

$2,152 → $2,179
BTC +6.2% same window

Thread · 1 event
  1. First reportAnthropic publishes new measurements tracking AI development and Claude’s AI R&D use40 acc.
  2. Leviathan published this article

All times UTC.

Cited posts

  • @AnthropicAI AnthropicFirst reportOn X ↗

    AI systems are getting more powerful, and they're increasingly being used to build the next version of themselves. We want to illuminate that progress for the public. Today, we're sharing three measurements that help track AI development: 1. How much AI R&D is done by AI. 2. How…

  • @FirstSquawk First SquawkOn X ↗

    ANTHROPIC SAID ITS MEASUREMENTS WILL FOCUS ON HOW AI MODELS ARE BUILT AND DEVELOPED, WITH THE DATA SHARED TO HELP TRACK PROGRESS; CLAUDE ACCOUNTED FOR 26% OF AI R&D USE AS OF AUGUST 2026.

  • @cyrilXBTOn X ↗

    Claude now leads 26% of Anthropic's own internal AI research, up from under 1% earlier this year. Roughly 30,000 agents are doing research and engineering work inside Anthropic at any single moment, and monitors block 1 in every 47,000 actions across a billion decisions. Only… https://t.co/iRKKsR2bo2

Show 3 more cited posts
  • @wallstengine Wall St EngineOn X ↗

    ANTHROPIC SAYS CLAUDE NOW “LEADS” 26% OF ITS OWN AI R&D Anthropic also disclosed that roughly 6% of its AI R&D compute goes to safety work, rising to about 12% for AI-driven AI R&D. As of August, Claude can complete most of the work end-to-end from a high-level prompt for 26%…

  • @upshot_cards UpshotOn X ↗

    JUST IN: Anthropic says Claude now leads 26% of its AI research and development. The share rose from about 1% in March. Over 90% of R&D now has AI collaborating or leading. About 30,000 agents run at once; every action is screened. Not fully autonomous yet.

  • @Prathkum PrathamOn X ↗

    February 2026: Claude led less than 1% of Anthropic's own AI R&D. August 2026: 26%. Claude now does over 90% of the work. And honestly it's a little terrifying that we are only seeing this now, because nobody outside these labs has had any way to check. – 30,000 agents are… https://t.co/jeDmT2duwt

34 more accounts posted about this · 1 Telegram channel (no public link)

TradersTrader calls tracked by Elfa · not advice

Positioning · ANTHROPIC

Open150 open positions · 106 long · 44 short

Tracked accounts lean long about 97% of the time, so a long-heavy book is normal, not a signal. As of 17 Sep 22:07 UTC.

Whole crypto market, not this story

Crowding66th percentile · vs last 28 days, 50 typical
Pulse 24h60/100 · unusualness of new calls, 0–100, 50 typical, higher means more long than usual
Book97.2% long · 12,003 accounts

As of 17 Sep 22:00 UTC.

Social-monitoring data, not investment advice Elfa prices Hyperliquid post text from Xas of

Explore the topic

Loading related coverage…

Explore the full feed

Comments

Loading comments…