Explore/benchmark/Persuasion Attacks Can Decrease Effectiveness of CoT Monitoring
P

Jennifer Za, Julija Bainiaksina, Nikita Ostrovsky, Tanush Chopra, Victoria Krakovna/Persuasion Attacks Can Decrease Effectiveness of CoT MonitoringUnknown

Chain-of-thought (CoT) monitoring is a promising safety mechanism for AI agents, based on the premise that visible reasoning traces can surface misaligned or deceptive behavior. While effective in standard scenarios, recent work highlights that LLMs remain vulnerable to persuasion-based jailbreaks, where natural-language arguments override model constraints. We stress-test whether this vulnerability extends to monitoring LLMs: can an adversarial agent persuade its CoT monitor to approve proposed actions that violate the monitor's policy? We design an evaluation framework with 40 tasks and analyze thousands of agent-monitor interactions, where agents are instructed to argue for policy-violating proposals. We find that in such adversarial settings, monitor access to the agent's CoT reasoning increases rather than decreases approval of harmful actions on average by 9.5%, as the scratchpad provides an additional persuasion channel. To address this, we introduce a fact-checking monitoring framework. We find that a fact-checker and monitor pairing from different model families, for example a Claude 3.7 Sonnet monitor paired with a GPT-4.1 fact-checker, reduces approval of policy-violating actions by up to 45%, compared to only 6%, when using the same model for both fact-checking and monitoring roles. Our results demonstrate that CoT monitoring alone may be insufficient against adversarial persuasion, and that model-diverse fact-checking provides a robust mitigation.

benchmark
GitHubCompare
Refreshed 2mo ago
OverviewActivity52wAlternativesDocs
Stars0
Forks0
HF Downloads—30d
Last commit—
Refreshed2mo ago
Project healthUnknownNo activity data.
Production readinessResearch / EarlyBest for exploration and prototyping.
Risk notesUnknown licenseVerify license before production use.
AgentHub Score
55 / 100
Composite score from 6 signals. How we score →
Active project
55Score
Growth
40C
Activity
30C
Documentation
70C+
Maturity
45C
Community
42C
Production
58C
GitHub stars · 79 days observed0 not enough history
snapshots
Repository activity · 79 days observedReal snapshots from pushed_at
inactivepushed
2026-07-262026-10-14
Practical assessment
Should you use it?

✓ Best for

  • Research and experimentation
  • Prototype development
  • Learning agentic patterns

◎ Strengths

  • Active community
  • Open source
  • Well-documented API

✕ Not ideal for

  • Untested at scale without validation
  • Teams without AI/ML expertise

⚠ Watch-outs

  • Review changelog before updating
  • Verify license for commercial use
Technical details
What's inside
Language—
License—
Sourcearxiv
Open source✗ No
Commercial use—
Docs—
Demo—

AgentHub Score

55
Score 55/100
Below average

Alternatives

R
RIFT-Bench: Dynamic Red-teaming For Agentic AI Systems
0 · benchmark
55
C
Commitment To Cooperation With Self-Negotiated Contracts
0 · benchmark
55
E
Exploring Structures in Physics Problems: Can AI Agents Discover Statistical Mechanical Mappings?
0 · benchmark
55
C
Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation
0 · benchmark
55Compare all →

Recent activity

Latest commit ——
Indexed by AgentHub crawler2mo ago
Monitor for new releasesongoing