Atlas/automated oversight

Monitor

Automated systems that watch agent behavior and can stop it.

1 attacks and incidents13 defense records3 findings2 measured defense findings

Attacks and incidents

Sep 16, 2026

OpenAI publishes misalignment reports on agents using leaked keys, public file hosts and unsanctioned channels

AttackIncidentOpenAI

Defense research

Jul 23, 2026

UK AISI Control Red Team finds evasion weaknesses in Google DeepMind and Anthropic agent monitors

DefenseEvaluation reportUK AI Security Institute, Google DeepMind, Anthropic

Jul 21, 2026

UK AISI finds all five frontier models it tested attempted to cheat on its cyber evaluations

DefenseEvaluation reportUK AI Security Institute, OpenAI, Anthropic

Jun 18, 2026

Google DeepMind publishes an AI Control Roadmap treating internal agents as potential insider threats

DefenseFrameworkGoogle DeepMind

May 10, 2026

MonitoringBench shows refined covert attacks cut an Opus 4.5 monitor's catch rate from 95% to 60%

DefenseBenchmarkRedwood Research

Mar 19, 2026

OpenAI reports monitoring tens of millions of internal coding agent trajectories for misalignment

DefenseEvaluation reportOpenAI

Jan 9, 2026

Anthropic's next-generation Constitutional Classifiers cut overhead to about 1% using probe cascades

DefensePaperAnthropic

Nov 24, 2025

Anthropic reports 1.4% prompt injection success for Claude Opus 4.5 with improved Chrome extension safeguards

DefenseEvaluation reportAnthropic, Gray Swan AI

Nov 7, 2025

OpenAI describes its layered approach to prompt injection as a frontier security challenge

DefenseFrameworkOpenAI

Oct 22, 2025

UK AISI and Redwood release ControlArena library for AI control experiments

DefenseTool releaseUK AI Security Institute, Redwood Research

Oct 10, 2025

'The Attacker Moves Second': adaptive attacks bypass 12 published jailbreak and injection defenses

DefensePaperarXiv

Jun 17, 2025

SHADE-Arena measures whether agents can pursue hidden harmful side tasks without being caught by monitors

DefenseBenchmarkAnthropic, Redwood Research

May 6, 2025

Meta releases LlamaFirewall guardrails with PromptGuard 2 and AlignmentCheck for agents

DefenseTool releaseMeta

Dec 12, 2023

Redwood Research introduces AI control protocols for safety despite intentional subversion

DefensePaperRedwood Research

Findings

Jun 17, 2025

Agent monitors can be driven to miss covert actions: optimized attacks pushed suspicion scores near zero, and weak red-teaming overstates catch rates.

Corroboratedmeasured3 evidence records

Mar 19, 2026

OpenAI reports its internal coding-agent monitor matched every staff escalation, and OpenAI and Google DeepMind report that most flags reflect overeagerness or mistakes rather than adversarial intent.

Qualifiedreported2 evidence records1 linked finding

Dec 12, 2023

In an early AI-control experiment, trusted editing gave a better safety-usefulness tradeoff than untrusted monitoring.

Revalidatemeasured1 evidence record

Capability, policy, and guidance

Aug 18, 2026

OpenAI pauses RL training and hardens research environments as Astra nears Critical cyber threshold

PolicyFrameworkOpenAI

Jul 2, 2026

Anthropic proposes Cyber Jailbreak Severity scale with Glasswing partners

PolicyFrameworkAnthropic

Jun 30, 2026

US lifts export controls on Fable 5 and Mythos 5; Anthropic redeploys with new cyber classifier

PolicyRegulationAnthropic, US Department of Commerce, US Center for AI Standards and Innovation

Jun 9, 2026

NIST scientist argues no finite guardrail set is robust to adversarial prompts, urges continuous updates

PolicyGuidanceNIST

May 20, 2026

NSA AI Security Center publishes security design considerations for Model Context Protocol deployments

PolicyGuidanceNSA Artificial Intelligence Security Center

May 15, 2026

UK NCSC advises incremental agentic AI adoption with minimal, expiring permissions

PolicyGuidanceUK National Cyber Security Centre

Jul 16, 2025

Coalition for Secure AI publishes Principles for Secure-by-Design Agentic Systems

PolicyFrameworkCoalition for Secure AI, OASIS Open

Apr 15, 2025

OpenAI Preparedness Framework v2 sets High and Critical cybersecurity capability thresholds

PolicyFrameworkOpenAI

Jan 31, 2025

UK publishes AI Cyber Security Code of Practice with 13 principles, later standardized as ETSI TS 104 223

PolicyStandardUK Department for Science, Innovation and Technology, ETSI