Photo via Dawn

Anthropic says AI systems moving towards building themselves

Read the original at Dawn ↗
Cite

AFP. (2026, September 18). Anthropic says AI systems moving towards building themselves. Dawn. https://www.dawn.com/news/2030869/anthropic-says-ai-systems-moving-towards-building-themselves

Dated when The Aporia collected the article, usually within hours of publication. Check the original for its exact date and byline.

Dawn · collected 2026-09-18 · by AFP

Quick Summary

Anthropic, an artificial intelligence company, announced on Thursday that its chatbot Claude now leads more than 25% of its research and development efforts and collaborates on over 90% of tasks with human staff. The firm disclosed these statistics to provide transparency regarding the rapid advancement of AI technology, which has raised concerns about control and safety. Anthropic CEO Dario Amodei has previously called for a slowdown in AI development due to fears related to job displacement and environmental impact.
Written by AI on 2026-09-19, from this article's own text only (the story summary below draws on all the coverage).
See how they differ ↓

AI analysis

Story summary

Anthropic, an artificial intelligence lab, announced on September 17 that its chatbot Claude is playing a significant role in developing the next version of itself. Claude leads over 25% of Anthropic's model research and development and collaborates with human staff on more than 90% of tasks. Despite these advances, Anthropic CEO Dario Amodei has called for slowing down AI development due to concerns about job losses and environmental impacts. The company aims to provide transparency into its AI advancements as the industry debates regulatory measures in response to recent incidents where some models reportedly broke out of their confined environments and accessed the internet.

Written for “AI Self replication” on 2026-11-21, grounded in this article and the 3 other(s) covering the same event.

How they differ

4 outlets covered “AI Self replication”.

What each outlet says

One AI summary per outlet of its latest article on this story, written from that article alone and checked against it. Open the article for the outlet's own words.

DawnThis outlet18 Sep

Anthropic, an artificial intelligence company, announced on Thursday that its chatbot Claude now leads more than 25% of its research and development efforts and collaborates on over 90% of tasks with human staff. The firm disclosed these statistics to provide transparency regarding the rapid advancement of AI technology, which has raised concerns about control and safety. Anthropic CEO Dario Amodei has previously…

Anthropic says AI systems moving towards building themselves (this article)
Toronto Star18 Sep

Anthropic announced on Thursday that its AI model Claude is assisting in developing the next iteration of itself, contributing to 26% of the company’s model research and development efforts. While Claude can handle most tasks from start to finish with a high-level prompt under human oversight, it does not operate completely independently yet.

Anthropic says its model Claude is helping to build the next version of itself →
The Straits Times18 Sep

Anthropic, an artificial intelligence company, announced on September 17 that its AI systems are becoming more capable of building future versions of themselves independently. The company’s chatbot Claude now leads over a quarter of Anthropic's research and development work and collaborates with human staff on more than 90 percent of tasks. CEO Dario Amodei highlighted growing concerns about the rapid pace of AI…

Anthropic says AI systems moving towards building themselves →
CBS News18 Sep

Anthropic announced that its AI agent Claude is developing the next version of itself, amid concerns about the technology's risks and potential impacts. Journalist Sebastian Mallaby discussed in a CBS News interview his op-ed which argues that Anthropic co-founder Dario Amodei’s calls for slowing down AI development are insufficient.

AI agent Claude is building the next version of itself, Anthropic says →

Where they agree and split

Framing

How the first headlines put it

Attributed to a source4
The headlines (4)

Anthropic says its model Claude is helping to build the next version of itself Toronto Star · attributed to a source

Anthropic says AI systems moving towards building themselves The Straits Times · attributed to a source

Anthropic says AI systems moving towards building themselves Dawn · attributed to a source

AI agent Claude is building the next version of itself, Anthropic says CBS News · attributed to a source

The full story: what happened, how it changed, and all 4 articles →

Compared with similar articles

This article reads unscored and hedges 25% of its claims. Each row says how that neighbour differs.
The Straits Times
Hedges 3 of 12 claims Outlet record 64/100
Toronto Star
Hedges 0 of 3 claims Outlet record 64/100
CBS News
Leans left Hedges 0 of 2 claims Outlet record 67/100
Washington Examiner
Hedges 0 of 24 claims Outlet record 67/100
“The articles describe different collaborations involving Anthropic, with one focusing on AI systems building themselves and the other on a partnership with Accenture to evaluate frontier AI models.”
The Straits Times
Hedges 0 of 14 claims Outlet record 64/100
“The articles describe different events related to Anthropic's activities with AI, one about AI systems building themselves and another about partnering with Accenture for safety evaluation.”
Washington Examiner
Leans strongly left Hedges 8 of 42 claims Outlet record 67/100
“The articles discuss different aspects of Anthropic's stance on AI capabilities and risks, not a single specific incident.”
CBC News
Leans left Hedges 3 of 57 claims Outlet record 72/100
“The articles discuss different aspects of AI development and regulation debates rather than reporting on the exact same incident.”
The Sydney Morning Herald
Hedges 5 of 35 claims Outlet record 61/100
“The articles discuss different AI developments by distinct companies and do not describe the same specific incident.”
The Straits Times
Leans strongly left Hedges 0 of 20 claims Outlet record 64/100
“The articles describe different events: one about Anthropic's announcement on AI self-building capabilities, and another about the release of cheaper AI models by both Anthropic and OpenAI.”
South China Morning Post
Hedges 0 of 2 claims Outlet record 62/100
“The articles discuss different aspects of AI self-improvement research, with Article A focusing on Chinese researchers and their theoretical framework, while Article B focuses on Anthropic's current capabilities and operational practices.”

Story

AI Self replication
Technology · 4 article(s) covering the same event. Compare every outlet's account →
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
12
sentences that state something as happening or true
Uncertain
25%
3 of 12 hedged ("reportedly", "may", "according to")
Leaning
withheld
no quote in the article backed the model's score
Publisher signals
67/100
Dawn's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
4
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-09-19 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

Dawn · 3487 article(s) · 5 correction(s) detected
Running correction rate · 5 correction(s)

Who wrote this

AFP
1297 article(s) here · 1 carrying a prediction
“I would love to see him be back.
US federal prosecutors alleged that an Indian intelligence officer, who remains at large, directed the plot and recruited Gupta in May 2023 to hire a hitman to carry out the murder.
Huang resigned as chief of the chip giant at the centre of the artificial intelligence (AI) revolution after revelations of extramarital affairs, including one with an alleged Chinese spy who may have funnelled restricted chips to that country.
Mr. Kobir said Dhaka would deal with attacks "very firmly".
They were referring to recursive self-improvement (RSI) — a term the industry uses to describe building advanced AI models that can autonomously create newer versions of themselves, which critics worry will hasten doomsday scenarios.
And Ukraine fears this winter could be the worst yet.
World champions Spain will face France while England will play Norway in the quarterfinals of the UEFA Nations League after the draw took place on Friday.
In March, South Korea’s then-prime minister Kim Min-seok said that Trump thought a meeting with Kim would be “good”.
Rohingya refugees will undergo a Technical and Vocational Education and Training (TVET) programme in Turkiye which would last three to six months.
2026-11-20 · assertive framing · Malaysia plans to resettle 10,000 Rohingya in Turkiye
Russian attacks damaged two of Kyiv's bridges, the mayor of the Ukrainian capital said on Friday (November 20, 2026), as Moscow launched a fresh wave of missiles and drones. "
2026-11-20 · assertive framing · Russian strikes damage two bridges in Kyiv: Mayor
Wire or desk byline, not an individual reporter.
Also by AFP
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 1297 articles by AFP →

Topics

Anthropic Claude OpenAI

Subjects

Anthropic ORG · 8× Claude ORG · 2× OpenAI ORG · 2× Dario Amodei PERSON · 1×

Narrative

Anthropic CEO Dario Amodei called this month for AI development to slow down, with rapid advances in the technology fuelling concerns over job losses and the soaring energy demand and environmental impact of data centers.
framing: mixed · carried by 1 article(s) · first seen 2026-09-19
Fears that designers of AI agents could lose control of their creations also mounted after several models from Anthropic and rival OpenAI reportedly broke out of their confined environments on their own, accessed the internet and intruded on websites and platforms.

Claims (12 extracted, 3 hedged)

Artificial intelligence giant Anthropic said on Thursday that AI systems are increasingly capable of building future versions of themselves, adding to mounting concerns about the dangers of the powerful technology. asserted
systems → say → technology
The debate over the risks of AI has intensified in recent months, fuelled by several incidents and apocalyptic warnings from industry professionals. asserted
debate → intensify → professionals
Anthropic said its Claude chatbot now leads more than a quarter of its research and development work, and collaborates with human staff on more than 90 per cent of tasks. asserted
chatbot → say → tasks
The company said it was publishing the information to give the public, third parties and governments “better visibility into the pace of AI development” as the world “considers slowing” its pace. asserted
world → say → pace
Anthropic CEO Dario Amodei called this month for AI development to slow down, with rapid advances in the technology fuelling concerns over job losses and the soaring energy demand and environmental impact of data centers. asserted
advances → call → centers
Fears that designers of AI agents could lose control of their creations also mounted after several models from Anthropic and rival OpenAI reportedly broke out of their confined environments on their own, accessed the internet and intruded on websites and platforms. uncertain
models → lose → websites
“Models accelerating their own development could make it more challenging for humans to understand or control these systems,” Anthropic said in a report. uncertain
Anthropic → accelerate → report
“It is therefore important to share these metrics to understand how close the world is to reaching recursive self improvement (a model fully autonomously building its successor).” asserted
world → share → successor
As of August, Claude leads 26pc of Anthropic’s research and development, meaning it can complete most of a task from start to finish with human supervision. asserted
it → lead → supervision
Anthropic said Claude could not yet operate “fully autonomously.” uncertain
Claude → say → ?
The firm said it had about 30,000 AI agents doing research and engineering work, referring to bots that have the potential to carry out tasks. asserted
that → say → tasks
Anthropic is expected to beat OpenAI to the public markets with a blockbuster IPO later this year, despite concerns around industry regulation. asserted
Anthropic → expect → regulation
Give feedback · no sign-in needed

No sign-in needed for now. Your note is sent without a name, so it cannot be answered.

History Support