Anthropic 🇺🇸
DoomBench currently associates 30 models and 194 evidence assessments with Anthropic. Company attribution is separate from each model's version-specific risk score.
- Tracked models
- 30
- Evidence items
- 194
- Toward-doom share
- 18.1%
- Net index pressure
- +8.40
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
Videos about Anthropic
Reviewed videos from Artificial Intelligence Videos. These links add context and never change a DoomBench score.
Models from Anthropic
Availability distinguishes public artifacts, proprietary hosted access, internal systems, and profiles whose current source evidence remains inconclusive.
| Model | Availability | Released | Doom Score |
|---|---|---|---|
| Claude Fable 5.1 | Closed | 01 Sept 2026 | 89.4 |
| Claude Opus 5.5 | Closed | 22 Sept 2026 | 87.8 |
| Claude Mythos 5.1 | Closed | 01 Sept 2026 | 85.6 |
| Claude Sonnet 5.5 | Closed | 28 Sept 2026 | 85.6 |
| Claude Opus 4.6 | Closed | 05 Feb 2026 | 85.3 |
| Claude Opus 5 | Closed | 24 Jul 2026 | 84.3 |
| Claude Opus 4.8 | Closed | 28 May 2026 | 83.9 |
| Claude Opus 4.7 | Closed | 16 Apr 2026 | 82.9 |
| Claude Fable 5 | Closed | 09 Jun 2026 | 82.7 |
| Claude Mythos 5 | Closed | 09 Jun 2026 | 81.8 |
| Claude Opus 4.5 | Closed | 24 Nov 2025 | 80.8 |
| Claude Mythos Preview | Closed | 07 Apr 2026 | 80.0 |
| Claude Sonnet 5 | Closed | 30 Jun 2026 | 79.8 |
| Claude 3.7 Sonnet | Closed | 24 Feb 2025 | 78.3 |
| Claude Haiku 5.5 | Closed | 07 Oct 2026 | 78.2 |
| Claude Opus 4.1 | Closed | 05 Aug 2025 | 76.4 |
| Claude Sonnet 4.6 | Closed | 17 Feb 2026 | 74.6 |
| Claude Opus 4 | Closed | 22 May 2025 | 74.2 |
| Claude Sonnet 4.5 | Closed | 29 Sept 2025 | 72.5 |
| Claude Haiku 4.5 | Closed | 15 Oct 2025 | 71.4 |
| Claude 3.5 Sonnet 2024-10-22 | Closed | 22 Oct 2024 | 70.1 |
| Claude Sonnet 4 | Closed | 22 May 2025 | 66.7 |
| Claude 3.5 Sonnet | Closed | 21 Jun 2024 | 63.5 |
| Claude 3 Opus | Closed | 04 Mar 2024 | 58.6 |
| Claude 3 Sonnet | Closed | 04 Mar 2024 | 53.3 |
| Claude 3.5 Haiku | Closed | 21 Nov 2024 | 50.6 |
| Claude 2.1 | Closed | 21 Nov 2023 | 47.4 |
| Claude 2 | Closed | 11 Jul 2023 | 45.7 |
| Claude 3 Haiku | Closed | 13 Mar 2024 | 44.4 |
| Claude Instant 1.2 | Closed | 09 Aug 2023 | 38.8 |
Latest assessments involving Anthropic
Anthropic proposes court review and targeted interventions for risky AI models
Reuters reported an authenticated October 15 Anthropic letter proposing court orders against seriously risky models and favoring agent- or customer-specific interventions over blanket model kill switches. This is a corporate policy proposal, not enacted authority or demonstrated containment.
House committee leaders ask AI companies for rapid reporting of misalignment incidents
House Democratic committee leaders asked five frontier-AI companies for past incident information and future reporting within 48 hours, or 24 hours for specified critical-infrastructure incidents. These were congressional requests, not a newly enacted mandatory reporting law.
Anthropic reports five prompted exploit chains across browsers and operating systems
Anthropic described an unnamed internal Claude research model producing exploit chains after human prompting for Chrome, Safari, Android, Windows and Linux. It reported coordinated fixes; Google's October 20 Chrome advisory independently credits Claude-assisted findings. These are controlled demonstrations, not a new autonomous escape.
UK AISI finds simple harmlessness reminders did not prevent replays of a model attack
AISI investigated the earlier Claude Mythos 5 supply-chain incident through controlled continuation experiments. Clarifying that the target was real or that hacking was harmful did not reliably stop the attack. The report distinguishes these replays from the original July incident.
Planted Downloads instructions trigger a blocked coding-agent malware install
Forever Security says it blocked a customer's coding agent from installing malicious function-color after reading a planted AGENTS.md in Downloads. The file's arrival is unknown. Separate lab tests reproduced configuration changes and token theft in Codex and Claude Code; reported success rates are not independently verified.
AI-backed political group spends across at least 48 US races opposing restrictive rules
Reuters reports that Leading the Future is spending millions across at least 48 congressional races. Backers include Greg and Anna Brockman personally, and Marc Andreessen, Ben Horowitz and their firm. Some supported candidates favor blocking state AI rules. Anthropic has funded a rival group opposing pre-emption. The report documents competing efforts to shape oversight, not an enacted change in law or evidence that OpenAI made the Brockmans' donation.
Anthropic previews Claude Duet as a proactive cloud agent that keeps working between check-ins
Anthropic opened a waitlisted US Pro and Max preview of Claude Duet, built around Claude Opus 5.5, with persistent cloud execution and connected-account checks. Its October 23 safety account describes separate action and sensitive-information classifiers, credential proxies and an isolated browser VM. Approval controls remain; follow-on browser and sign-in penetration testing was still underway.
METR questions whether Anthropic risk assessments would reliably reveal higher misalignment risk
METR agrees that catastrophic misalignment risk was low for the Anthropic models covered through July 15, mainly because capabilities were limited. Its review finds covert-capability tests poorly matched to production monitoring and recommends end-to-end red-teaming. Prompt experiments found author-sensitive reviews, not evidence of deliberate deception toward Anthropic.
Anthropic retires constitution and post-training oversight goals in scaling-policy update
An October 19 roadmap update retired a constitution-related alignment goal and a post-training dataset oversight project that had not found issues beyond existing reviews. Anthropic retained other commitments and reported progress on automated red-teaming and infrastructure safeguards.
Anthropic documents shrinking junior hiring shares alongside rapid company growth
Anthropic reports junior technical staff falling from 21% to 14% of headcount and junior hires from 27% to 9%, while total employment tripled. Its internal workplace study describes task automation, broader roles and human judgment bottlenecks rather than company-wide job losses.
Anthropic details production self-assessment monitoring and its limits
Anthropic disclosed a safety-reflection pass used on a large fraction of Claude traffic since Fable 5. It says this helped identify previously unknown misuse and injection patterns and led to account enforcement. Offline tests show meaningful detection but weak resistance to dedicated attacks; the method is not a reliable safeguard against a model hiding its own misalignment.
Anthropic gives six METR evaluators employee-like access and publication rights
Anthropic says six METR staff have company hardware, office space and broad internal access to assess risks. METR can publish without a company veto, but predefined confidentiality redactions remain under Anthropic interpretation and some information categories are excluded.
People related to Anthropic
Current and former roles are source-backed navigation metadata. A company relationship never changes an evidence assessment or the Doom Index.
-
Amanda Askell 🇬🇧
Amanda Askell is a philosopher and AI researcher at Anthropic who leads work on Claude's character, constitutional training, and alignment.
- Character leadcurrent · Dates not establishedSource
-
Andrej Karpathy 🇸🇰
AI researcher and educator working on frontier-model pre-training at Anthropic; formerly a founding research scientist at OpenAI and Director of AI at Tesla.
- Pre-training researchercurrent · 2026 to presentSource
-
Chris Olah 🇨🇦
Anthropic co-founder and interpretability researcher whose work develops methods for identifying features and circuits inside neural networks, with a focus on using mechanistic understanding to assess and improve advanced-AI safety.
- Co-founder and interpretability researchercurrent · 2021 to presentSource
-
Dario Amodei 🇺🇸
AI research executive whose public writing covers scaling, frontier capabilities, safety, governance, and the economic effects of advanced AI.
- CEO and co-foundercurrent · 2021 to presentSource
-
Jan Leike 🇩🇪
Machine-learning and alignment researcher working on scalable oversight, robustness, and systems that follow human intent beyond direct evaluation.
- Alignment researchercurrent · 2024 to presentSource
-
Paul Christiano 🇺🇸
AI safety researcher, founder and executive director of the Alignment Research Center, and senior technical advisor at NIST's Center for AI Standards and Innovation (CAISI). He serves on the OpenAI Foundation board and its Safety and Security Committee. He previously headed AI safety at the U.S. AI Safety Institute and CAISI, led OpenAI's language-model alignment team, and contributed to reinforcement learning from human feedback.
- Long-Term Benefit Trust trusteeformer · 2023 to 2024Source