Anthropic announces Claude assists in building its next AI models and partners with Accenture for embedded evaluation.
Anthropic disclosed that its Claude model now leads approximately 26% of the company's research and development work, with about 90% done in collaboration with Claude. The company also announced a partnership with Accenture to independently evaluate frontier AI models, with each investing at least $1 billion over five years. These announcements follow an Anthropic researcher's resignation citing existential AI risks.
47
Divergence score
This event sits in the top 5% of divergence this week. 15 outlets covered it, splitting into 15 framing camps across 4 bias groups.
15 camps
4 bias groups
Market signalBETA
The spectrum · how 15 outlets placed this story
LeftCenterRight
Washington Examiner
Washington Times
Financial Times
Wall Street Journal
Reuters
The Hill
Globe and Mail
Al Jazeera
Breitbart
Jerusalem Post
CNN
South China Morning Post
BBC
Axios
NY Post
Supportive of action
Neutral
Dismissive
Critical
Alarmist
International angle
Widest split · the two outlets furthest apart, and who published first
Furthest apart on this story
Washington Examiner
Anthropic partners with IT company Accenture to evaluate frontier AI models
safety governance partnership · neutral
CNN
Anthropic says its AI models pose ‘existential risk to humanity’ in leaked IPO filing: report - CNN
Existential risk disclosure amid IPO valuation · negative
First to report
- Reutersfirst
- Wall Street Journal+1.0 h
- Washington Times+1.9 h
- The Hill+18.4 h
- Washington Examiner+22.9 h
- Financial Times+23.7 h
- Globe and Mail+117.2 h
- Al Jazeera+148.4 h
- Breitbart+160.5 h
- Jerusalem Post+272.2 h
- CNN+278.7 h
- South China Morning Post+300.0 h
By each outlet's earliest publish time in this story. Feeds that index slowly can appear later than they printed.
Headlines that changed · the same article, re-listed with a new headline
Wall Street Journal · 2026-09-18 03:35 UTC
Exclusive | Hackers Used Anthropic’s Claude to Break Into OpenAI - wsj.com
Reuters · 2026-09-18 06:58 UTC
Anthropic says Claude now leads a quarter of work building its next AI models - Reuters
Reuters · 2026-09-18 07:31 UTC
Anthropic says Claude now leads a quarter of work building its next AI models - reuters.com
Reuters · 2026-09-18 08:05 UTC
Anthropic says Claude now leads a quarter of work building its next AI models - Reuters
Financial Times · 2026-09-19 12:53 UTC
Anthropic brings in Accenture for AI safety testing - Financial Times
+ Show 8 more Show fewer
Financial Times · 2026-09-19 13:26 UTC
Anthropic brings in Accenture for AI safety testing - ft.com
Financial Times · 2026-09-19 22:41 UTC
Anthropic brings in Accenture for AI safety testing - Financial Times
Financial Times · 2026-09-19 23:16 UTC
Anthropic brings in Accenture for AI safety testing - ft.com
Financial Times · 2026-09-20 04:34 UTC
Anthropic brings in Accenture for AI safety testing - Financial Times
Financial Times · 2026-09-20 05:09 UTC
Anthropic brings in Accenture for AI safety testing - ft.com
Financial Times · 2026-09-20 14:47 UTC
Anthropic brings in Accenture for AI safety testing - Financial Times
CNN · 2026-09-29 21:24 UTC
Anthropic says its AI models pose ‘existential risk to humanity’ in leaked IPO filing: report | CNN Business - CNN
BBC · 2026-10-05 17:12 UTC
Pentagon stops using Anthropic AI tools months after blacklisting company, BBC told
The split, in one line
Coverage now splits between Anthropic's proactive critical infrastructure defense and Pentagon skepticism over supply chain control, while ethical guardrails emerge as a separate debate, exposing tension between industry self-governance, government oversight demands, and anthropomorphic user protections amid asymmetric competition with Chinese AI.
How each outlet covered it
The split runs within one side
The sharpest contrast here isn't left vs right: it's between outlets that usually agree.
FRAMED AS A PROBLEM
“Anthropic says Claude is helping to build the next version of itself”WT Washington Times RIGHT
47DIVERGENCE
FRAMED AS PRAGMATIC
“Mad Scientists: Anthropic Claims Claude AI Has Discovered an Enzyme with Gene-Editing Potential”B Breitbart RIGHT
DOWN THE MIDDLE
“Anthropic says Claude now leads a quarter of work building its next AI models - reuters.com” · Reuters, The Hill, Globe and Mail, Al Jazeera, Jerusalem Post, South China Morning Post, BBC, Axios
+Hide the full sourcingSee how all 15 outlets put it
CENTER3
RReuters Anthropic says Claude now leads a quarter of work building its next AI models - reuters.com 45d ago HThe Hill Anthropic: AI model Claude helping build next version of it 44d ago AAxios Exclusive: Anthropic's plan to protect critical infrastructure 24d ago RIGHT-CENTER2
FTFinancial Times Anthropic brings in Accenture for AI safety testing - Financial Times 44d ago WSJWall Street Journal Exclusive | Hackers Used Anthropic's Claude to Break Into OpenAI - WSJ 45d ago RIGHT4
WEWashington Examiner Anthropic partners with IT company Accenture to evaluate frontier AI models 44d ago WTWashington Times Anthropic says Claude is helping to build the next version of itself 45d ago BBreitbart Mad Scientists: Anthropic Claims Claude AI Has Discovered an Enzyme with Gene-Editing Potential 38d ago NYPNY Post Anthropic to users: be nice to Claude AI — new policy bars ‘abusive or cruel behavior’ to tech 24d ago INTERNATIONAL5
GMGlobe and Mail Anthropic unveils cheaper AI model Claude Opus 5.5 40d ago AJAl Jazeera AI model Claude discovers CRISPR-like enzyme system, Anthropic says 39d ago JPJerusalem Post 'Virtuous' AI company: Anthropic leaders control AI lab via new 'Founder LLC' to serve public good 33d ago SCSouth China Morning Post Anthropic raises alarm over elite hacking ability of Chinese firm Z.ai’s GLM-5.3 32d ago BBCBBC Pentagon stops using Anthropic AI tools months after blacklisting company, BBC told 27d ago 6 tracked claims across 15 outlets
Fact ledger
All6Claimed1Corroborated5
1/15
Claimed
Hackers used Anthropic's Claude to break into OpenAI.
Corroborated
Disputed
5 more tracked claimsSign up free to see the full ledger