Anthropic Confirms Testing Powerful Claude Mythos AI Model, Citing Release Dangers

A data leak and subsequent reports revealed details about Anthropic's latest secret AI model, named Claude Mythos or a 'Step Change' model, which has led to a slump in cybersecurity stocks. Anthropic has now confirmed it is testing this new AI model, describing it as their most capable yet, but also the most dangerous to release, introducing a new tier above Opus called Capybara.

27 Mar, 12:50 — 27 Mar, 13:01

Post Share

Why this matters

The development of increasingly powerful AI models like 'Claude Mythos' raises significant questions about safety, ethical deployment, and the potential for unforeseen societal impacts. Anthropic's decision to withhold such a model highlights the growing tension between rapid technological advancement and the need for responsible innovation in the AI field.

Key entitiesAnthropic Claude Mythos

The Story

What 4 sources agree on, dispute, and miss

What sources agree on

Anthropic is testing a new, powerful AI model named 'Claude Mythos'
The existence of 'Claude Mythos' was revealed through an accidental data leak
Anthropic considers 'Claude Mythos' to be its most capable model to date
Anthropic has expressed concerns about the dangers of releasing 'Claude Mythos' publicly

Key claims4 agreed · 1 unverified

✓

Anthropic is testing an AI model named 'Claude Mythos'

agreed·le-figaroseeking-alphazerohedgeTimes of India

✓

The model's existence was revealed through an accidental data leak

agreed·le-figaroseeking-alphazerohedge

✓

Anthropic considers 'Claude Mythos' its most powerful AI model

agreed·le-figaroTimes of India

✓

Anthropic is refusing to release 'Claude Mythos' due to safety concerns

agreed·le-figaroTimes of India

Cybersecurity stocks tumbled in premarket trading following the report of the new AI model

unverified·zerohedge

Coverage gaps

The impact of the news on cybersecurity stocks, which reportedly tumbled in premarket trading

Reportedzerohedge

Missingle-figaroseeking-alphaTimes of India

Specific details about the 'new tier above Opus' that Claude Mythos introduces

ReportedTimes of India

Missingle-figaroseeking-alphazerohedge

The Story

Analyzing sources…

Source Diversity

High (51/100)

4 sources30/33

Spectrum spread— 2/5 buckets covered8/33

Far L

Left

Center

Right3

Right (3)

le-figaroseeking-alphaTimes of India

Far R1

Far Right (1)

zerohedge

Geographic diversity— 3 regions13/34

US2France1India1

Sources

Showing 2 of 4 sources

Times of IndiaMostly Factual11h ago

Why Anthropic is 'refusing' to release AI model that it says is most powerful it made

Anthropic has confirmed it is testing Claude Mythos, a new AI model it describes as the most capable it has ever built—and the most dangerous to release. The model introduces a new tier above Opus called Capybara, and significantly outperforms Claude Opus 4.6 in coding, reasoning, and cybersecurity. Anthropic isn't rushing a wider rollout because, by its own assessment, Mythos is currently far ahead of every other AI model in cyber capabilities.

By TOI TECH DESK

Read full article →

zerohedgeLow11h ago

AI Disruption Returns: Cybersecurity Stocks Tumble On Report Of New Anthropic "Step Change" AI Model

AI Disruption Returns: Cybersecurity Stocks Tumble On Report Of New Anthropic "Step Change" AI Model Cybersecurity stocks are slumping in premarket trading on Friday as the AI disruption trade returns after a Fortune report said Anthropic is testing a new AI model that “poses significant cybersecurity risks.” According to the report, Anthropic is "developing and has begun testing with early access customers a new AI model more capable than any it has released previously," the co...

By Tyler Durden

Read full article →

Coverage Timeline

First report: Times of India · 27 Mar, 12:50|Full coverage: 4 · 6h|Window: 6h

Left-leaningCenterRight-leaning

Times of India27 Mar, 12:50First to report

Why Anthropic is 'refusing' to release AI model that it says is most powerful it made

zerohedge27 Mar, 13:01

AI Disruption Returns: Cybersecurity Stocks Tumble On Report Of New Anthropic "Step Change" AI Model

seeking-alpha27 Mar, 13:30

Data leak reveals Anthropic's latest secret model, Claude Mythos: report

6h later

le-figaro27 Mar, 19:08Latest update

«Trop puissant» pour une diffusion publique : le prochain modèle d’IA d’Anthropic, victime d’une fuite, suscite la peur de ses créateurs

Story evolution

The initial narrative focused on the leak and market impact, shifting to Anthropic's official confirmation and their explicit concerns about the model's power and potential dangers.

Initial Leak and Market ReactionMarch 27 morning

Reports emerged about an accidental data leak revealing Anthropic's new 'Claude Mythos' model, leading to a reported tumble in cybersecurity stocks.

Anthropic Confirmation and Danger AssessmentMarch 27 afternoon

Anthropic confirmed the existence of 'Claude Mythos,' describing it as their most capable model and expressing concerns about its public release due to potential dangers.

PERSPECTA

PERSPECTA

Anthropic Confirms Testing Powerful Claude Mythos AI Model, Citing Release Dangers

Why this matters

The Story

What sources agree on

Key claims4 agreed · 1 unverified

Coverage gaps

The Story

Source Diversity

Source Diversity

Sources

Why Anthropic is 'refusing' to release AI model that it says is most powerful it made

AI Disruption Returns: Cybersecurity Stocks Tumble On Report Of New Anthropic "Step Change" AI Model

Coverage Timeline

Why Anthropic is 'refusing' to release AI model that it says is most powerful it made

AI Disruption Returns: Cybersecurity Stocks Tumble On Report Of New Anthropic "Step Change" AI Model

Data leak reveals Anthropic's latest secret model, Claude Mythos: report

«Trop puissant» pour une diffusion publique : le prochain modèle d’IA d’Anthropic, victime d’une fuite, suscite la peur de ses créateurs

Story evolution

Related Stories

NASA Astronaut Mike Fincke's Mysterious Space Illness Led to Mission Cancellation

Philippine Customs Bureau Proposes Costly ITAMSS System Despite Free Alternative

Washington D.C. Area Airports Face Ground Stops Due to Odor at Air Traffic Control Tower

Portuguese Schools to Be Rebuilt to Withstand Extreme Weather After Kristin

PERSPECTA

Why this matters

Story evolution

The Story

What sources agree on

Key claims4 agreed · 1 unverified

Coverage gaps

Coverage matrix