Skip to content
Today
Sections

Invest Independently confirmed4 publishers 3 min read Published Updated

Anthropic invites its rivals to benchmark against a self-reported 26%

Anthropic says Claude now leads 26% of its model research and development under human supervision, and it wants other labs to publish comparable figures. The labs count different things.

The Investor · Invest desk

Send a correction

What happened

  • The disclosure gave rival labs a look at Anthropic's progress toward recursive self-improvement, and the company encouraged those competitors to publish similar metrics of their own.
  • OpenAI said this month it had built an automated "research intern" that handles well-defined research tasks under human direction, including tasks that would take a skilled researcher a few days.
  • Fortune reports that leading AI companies define recursive self-improvement differently, some counting any AI feedback on model improvement and others only fully autonomous self-design.

Why it matters

  • What doesn't add up contradiction Anthropic wants comparable numbers from labs that do not agree on what to count, so a 26% from one and a research intern from another cannot be put in the same column.
  • Who has to decide decision Every rival lab now chooses between publishing in Anthropic's unit, publishing in its own, or staying quiet, and the second publisher effectively ratifies or rejects Anthropic's definition.
  • What it sets up precedent OpenAI has attached a calendar date, March 2028, to an automated researcher. A date is something outsiders can hold a lab to.
  • What it limits constraint The missing denominator leaves two readings open: work taken off researchers' desks, or work reclassified into a category Claude already handles.

Next story · Invest · Sep 21

Four paid subscribers turn the AI pacing essays into a Sherman Act complaint

Confirmed2 publishers

"Leading" 26% of model research and development means, in Anthropic's own gloss, that Claude can complete most of a given task end-to-end from a high-level prompt while a human still supervises [2]. So the figure sorts tasks by how much prompting they need, and it leaves 74% of that research and development not led by Claude [20]. Fortune's account explains what leading means; it does not report how Anthropic defines the total the 26% is a share of [7].

The invitation to competitors is hard to accept on its own terms [4]. Fortune reports that the leading labs do not agree on what recursive self-improvement is, with some counting any feedback from AI on model improvement and others only a system working toward it fully autonomously [6]. Under the loose version, the number would have been sizeable years ago. John Thickstun, a Cornell assistant professor of computer science who studies methods that control model behaviour, said: "We have already, for years, been using these models in supportive roles for creating the next version of these models. So people use the past generation of models to write code for the AI systems that then create the next generation" [11]. He said the older attempts, including those by OpenAI co-founder Andrej Karpathy, produced minor improvements and not big creative leaps [12].

OpenAI answered in a different unit. It said this month it had built an automated "research intern" that carries out well-defined research tasks under human direction, including "tasks that would take a skilled researcher a few days" [13], and it has put a date on the full automated "researcher": March 2028 [14], about 18 months out [21]. It also wrote in that announcement that "Whether and how to proceed must depend on our ability to preserve human control and on informed democratic choices about the benefits and risks" [8]. Elon Musk said in March that for xAI's Grok models, "humans are gradually getting less and less in the loop" [15].

In my view the disclosure practice is the more interesting part of this, or rather, the sequencing is. Anthropic published a figure and asked rivals to publish theirs [4], and whoever goes second either adopts Anthropic's definition of a task or explains why it prefers its own. Anthony Aguirre, president and CEO of the Future of Life Institute and a physics professor at UC Santa Cruz, said "You can see in these plots from Anthropic over time, more and more of research is being done by the AI and it's becoming closer and closer to fully autonomous" [16], and "I think this is probably the worst idea in the history of humanity to do this. And yes, they're doing it" [17].

What would settle it: a second Anthropic disclosure that pairs the percentage with a denominator and a research headcount or spend line that has not grown, which would make 26% a productivity figure with money behind it. If the percentage climbs while the definition moves, it is positioning. Anthropic itself has stayed quiet on how close it is to fully autonomous model improvement [5].

What to watch

  • Whether any competing lab publishes a figure in Anthropic's unit, and whether it supplies the denominator Anthropic left out.
  • Whether Anthropic's next update moves the percentage, the definition of a task, or both.
  • Whether OpenAI restates its March 2028 automated researcher target as a percentage of work completed, or lets the date slip.

Information, not investment advice. This story is general information about companies and markets, drawn from cited reporting. It is not a recommendation to buy, sell or hold any security; check the cited sources before acting.

Clarity's read

What the record supports and how the coverage leans. The claims behind it follow.

Reality

Evidence56
Adoption52
Hype gap+24
Incentives72
Confidence66

Perspective Coverage

4 publishers
Builder
Builder 23%
Operator
Operator 41%
Investor
Investor 36%
Why these scores

Claim ledger

Ranked by verification strength, evidence, and original report placement.

  1. [1]

    Anthropic this week detailed how its model Claude is helping the company develop the next, more intelligent version of itself.

    Reported SupportedView cited source
  2. [2]

    Claude is now leading 26% of Anthropic's model research and development, which the company said means it can complete most of a given task "end-to-end from a high-level prompt" while still being under human supervision.

    Reported SupportedView cited source
  3. [3]

    Fortune reports that the models are not working completely autonomously, at least not yet.

    Reported SupportedView cited source
  4. [4]

    Anthropic's announcement gave the public and other labs insight into its RSI progress, and it encouraged its competitors to share similar metrics.

    Reported SupportedView cited source
  5. [5]

    Anthropic has not expressly said how close it is to achieving fully autonomous model improvement.

    Reported SupportedView cited source
  6. [6]

    Leading AI companies have different definitions for recursive self-improvement: some define it as any feedback from AI on model improvement, while others define it as AI working toward that goal fully autonomously.

    Reported SupportedView cited source
  7. [7]

    Fortune's account states what Anthropic means by Claude "leading" a task but does not report how Anthropic defines the total volume of model research and development against which the 26% is measured.

    Reported SupportedView cited source
  8. [8]

    OpenAI said in a blog post: "Whether and how to proceed must depend on our ability to preserve human control and on informed democratic choices about the benefits and risks."

    Reported Supported Source: OpenAI blog post, quoted by FortuneView cited source
  9. [9]

    OpenAI said that while RSI can help align models' actions with human values and intentions, that does not mean "rapid RSI is necessarily an outcome we should pursue".

    Reported SupportedView cited source
  10. [10]

    Uncertainty over where recursive self-improvement leads led several AI moguls to join last weekend in a call to slow down the technology's pace of growth.

    Reported SupportedView cited source
  11. [11]

    John Thickstun, an assistant professor of computer science at Cornell University who studies methods that control the behavior of AI models, said: "We have already, for years, been using these models in supportive roles for creating the next version of these models. So people use the past generation of models to write code for the AI systems that then create the next generation."

    Reported Supported Source: John Thickstun, Cornell University, quoted by FortuneView cited source
  12. [12]

    Thickstun said that years of experiments by prominent AI researchers such as OpenAI co-founder Andrej Karpathy, trying to get AI models to train and improve new AI systems, brought minor improvements but not big creative leaps.

    Reported SupportedView cited source
  13. [13]

    OpenAI announced this month that it has developed an automated "research intern", which it defines as a system that can carry out well-defined research tasks under human direction, including "tasks that would take a skilled researcher a few days".

    Reported SupportedView cited source
  14. [14]

    OpenAI has said it is moving forward with the goal of creating an automated AI "researcher" by March 2028.

    Reported SupportedView cited source
  15. [15]

    Elon Musk said in March that for xAI's Grok models, "humans are gradually getting less and less in the loop".

    Reported Supported Source: Elon Musk, quoted by FortuneView cited source
  16. [16]

    Anthony Aguirre, president and CEO of the nonprofit Future of Life Institute and a physics professor at the University of California, Santa Cruz, said: "You can see in these plots from Anthropic over time, more and more of research is being done by the AI and it's becoming closer and closer to fully autonomous."

    Reported Supported Source: Anthony Aguirre, Future of Life Institute, quoted by FortuneView cited source
  17. [17]

    Aguirre said: "And the result of that success, ultimately is something that is, I think, extremely scary. I think this is probably the worst idea in the history of humanity to do this. And yes, they're doing it."

    Reported Supported Source: Anthony Aguirre, quoted by FortuneView cited source
  18. [18]

    Aguirre said: "The really important thing here is that as AI is doing more of it, it gets faster, because AI operates just much, much more quickly than the humans do."

    Reported Supported Source: Anthony Aguirre, quoted by FortuneView cited source
  19. [19]

    Fortune published its account of recursive self-improvement on September 19, 2026.

    Reported SupportedView cited source
  20. [20]

    If Claude leads 26% of Anthropic's model research and development, 74% of it is not led by Claude.

    Derived Supported
  21. [21]

    OpenAI's March 2028 target for an automated AI researcher is about 18 months after the September 2026 report of it.

    Derived Supported

Sources

4 independent publishers whose own reporting we read for this story.

  1. bloomberg.com

    1 article · September 17, 2026

    Anthropic Says Claude Drives 26% of Its Research and Development - Bloomberg
  2. fortune.com

    1 article · September 19, 2026

    As AI companies get closer to ‘recursive self-improvement,’ this physics professor calls full autonomy ‘the worst idea in the history of humanity’
  3. livemint.com

    1 article · September 17, 2026

    Anthropic says Claude now leads a quarter of work building its next AI models | Mint
  4. nbcnews.com

    1 article · September 18, 2026

    Anthropic says its model Claude is helping to build the next version of itself

Topics and entities

Topics

Entities