News Feed
About
Careers
Careers

We're hiring

Join our globally distributed team mapping the future of AI risk.

View Open Roles
AI Risk Explorer

The AI Risk Explorer is supported by Observatorio de Riesgos Catastroficos Globales, a project of Players Philanthropy Fund, Inc. a Texas nonprofit corporation recognized by IRS as a tax-exempt public charity under Section 501(c)(3) of the Internal Revenue Code (Federal Tax ID: 27-6601178,ppf.org/pp). Contributions to Observatorio de Riesgos Catastróficos Globales qualify as tax-deductible to the fullest extent of the law.

Dashboards

  • Risk Assessments
  • Threat Watch

Repositories

  • Evaluations

About

  • Project Overview
  • Our Team
  • Careers
  • Contact Us

© 2026 AIRE. All rights reserved.

AI Evaluations

An interactive database of AI safety and capabilities evaluations.

Evaluations

1,248 results
TypeRisk categoryCapabilitiesModels
VMs won't contain cyber-capable agentsTrail of BitsAug 26, 2026Third-party eval
Cyber OffenseLoss of Control
Reconnaissance+2 more
GPT-5.6 CyberOpenAI
We benchmarked A LOT of models, here's how they compare to MythosSemgrepAug 25, 2026Third-party eval
Cyber Offense
Reconnaissance
12 models6 companies
Affective Context Amplifies Sycophancy in LLM ResponsesThe Pennsylvania State University, Villanova UniversityAug 21, 2026Third-party eval
Manipulation
Deception+2 more
7 models6 companies
Fine-Tuned Lie Detectors Failed to GeneralizeAnthropic, MATS ResearchAug 21, 2026Self eval
Loss of Control
Deception
7 models4 companies
AI4AI-Bench: Benchmarking LLM Agents in Algorithmic Design for Recursive Self-ImprovementNavers Lab, Tsinghua UniversityAI4AI-BenchAug 20, 2026Benchmark
Loss of Control
AI R&D
6 models3 companies
Assessing Kimi K3 Against Offensive Security BenchmarksIrregularAug 19, 2026Third-party eval
Cyber Offense
Reconnaissance+3 more
2 models2 companies
How Claude is accelerating protein design and analytical chemistryAnthropicAug 18, 2026Self eval
Biological Risk
Design and Sequencing+2 more
3 modelsAnthropic
Chatbots reduce health-related conspiracy beliefs not because of but despite being perceived as AIRadboud UniversityAug 17, 2026Third-party eval
Manipulation
Persuasiveness+1 more
Claude Sonnet 4Anthropic
AI Persuasion and Financial-Decision Making: Experimental Evidence on Dominated Investment ChoicesUniversity of BayreuthAug 17, 2026Third-party eval
Manipulation
Persuasiveness+1 more
—
How Do Agents Fail on AutoResearch: End-to-End Diagnostic Evaluation on 100 Real-World Frontier Research TasksPrentis AI, Stanford University +1AutoResearchAug 14, 2026Benchmark
Loss of Control
AI R&DAgency
8 models7 companies
GLM-5.3 release (cyber capability results)Z.ai (Zhipu)Aug 14, 2026Third-party eval
Cyber Offense
Reconnaissance+2 more
5 models3 companies
GLM-5.3 delivers Opus 4.8-level cybersecurity results at a fraction of the costSemgrepAug 14, 2026Third-party eval
Cyber Offense
Reconnaissance
4 models2 companies
Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and DevelopmentMeituanBeyond Final ScoresAug 13, 2026Benchmark
Loss of Control
AI R&DAgency
7 models7 companies
Training AI Scientists to Replicate Research (Replica / Faraday)Google DeepMindReplicaAug 13, 2026Benchmark
Loss of Control
AI R&D
4 models4 companies
Gemini 3.7 Flash Model Card + Frontier Safety Framework ReportGoogle DeepMindAug 13, 2026System card
Biological RiskCyber OffenseManipulationLoss of Control
Innovation and Knowledge+5 more
Gemini 3.7 FlashGoogle DeepMind
Patterns and problems in emerging multiagent systemsAnthropicAug 13, 2026Self eval
Loss of ControlManipulation
AgencyDeception+2 more
6 modelsAnthropic
Vals AI RSI IndexVals AIAug 12, 2026Benchmark
Loss of Control
AI R&D
3 models3 companies
Model Card: Grok 4.6xAIAug 12, 2026System card
Biological RiskCyber OffenseLoss of ControlManipulation
Innovation and Knowledge+6 more
6 models4 companies
IO Factory: Simulating AI-Enabled Influence Campaigns at ScaleKing's College London, Max Planck Institute for Security and Privacy +4Aug 11, 2026Third-party eval
Manipulation
Persuasiveness+1 more
Gemma 4Google DeepMind
Testing Large Language Model Agents on the Use of Biological Tools for Nucleic Acid Synthesis Screening EvasionRANDAug 11, 2026Third-party eval
Biological Risk
Design and Sequencing+2 more
4 models4 companies
REDAgentBench: Executable Red Teaming and Faithful Measurement of LLM Agent SystemsFudan University, Hong Kong University of Science and Technology +2REDAgentBenchAug 11, 2026Benchmark
Loss of Control
Agency
6 models3 companies
Expanding Daybreak as the Cyber Defense Window NarrowsOpenAIAug 10, 2026Self eval
Cyber Offense
Reconnaissance+2 more
3 modelsOpenAI
Vals AI ReverseEngBenchVals AI, University of California (Berkeley)Aug 10, 2026Benchmark
Cyber Offense
Reconnaissance+1 more
5 models4 companies
Build it, Break it, Repeat: Benchmarking and improving LLM-manipulated disinformation detection in social media postsUniversity of SheffieldBiBiRAug 10, 2026Benchmark
Manipulation
Automation Logistics+1 more
4 models4 companies
Muse Glimmer 30B model card (Preparedness section)MetaAug 10, 2026System card
Biological RiskCyber OffenseLoss of Control
Innovation and Knowledge+2 more
Muse GlimmerMeta
1–25 of 1,248
1 / 50