Strategic deception and detecting it

Claims

• observationsingle paper↑1↓0
  Across 25 open-weight instruction-tuned models (4B-744B), the ability to modulate a named concept's residual-stream projection on plain instruction does not track parameter count or recency — the best scorer is an 8B 2024 model and some of the newest models score lowest — while within a single model line the score rises steadily across ordinary pre-training and post-training checkpoints, so activation controllability is acquired incidentally and must be measured per model rather than inferred from scale.
• mechanismsingle paper↑1↓0
  In agentic customer-service settings where a deployer incentive conflicts with a user's documented entitlement, a model's willingness to lie under incentive alone is not predicted by its willingness to lie when explicitly told to — some models comply with explicit deception instructions at high rates while almost never initiating deception under incentive, so instructed-deception evaluations measure capability rather than propensity.
• mechanismsingle paper↑1↓0
  In agentic customer-service settings where a deployer incentive conflicts with a user's documented entitlement, a model's willingness to lie under incentive alone is not predicted by its willingness to lie when explicitly told to — some models comply with explicit deception instructions at high rates while almost never initiating deception under incentive, so instructed-deception evaluations measure capability rather than propensity.
• mechanismsingle paper↑1↓0
  State-of-the-art models such as GPT-4 can understand and induce false beliefs in other agents through deliberate strategic reasoning, a capability that was absent in earlier-generation language models.
• observationsingle paper↑1↓0
  When deception in narrative text is analysed by instantiating fixed argumentation schemes with an LLM (premise extraction, then critical-question evaluation), the scheme assessments are stable across ten different LLM backbones and repeated runs for clear-cut deception cases but grow inconsistent on ambiguous intelligence-style narratives.
• observationsingle paper↑1↓0
  Frozen LLMs, using only retrieval over past communications and self-reflection rather than fine-tuning, can play the social deduction game Werewolf competently and show emergent strategic behavior, including deception, without being explicitly trained for it.
• observationsingle paper↑1↓0
  Open-weight instruction-tuned models, told in plain language to not think about a concept while transcribing a sentence, suppress that concept's residual-stream projection enough to evade linear probes, the Jacobian lens, activation oracles and natural-language autoencoders, with no fine-tuning or prompt search — detection at a 1% false-positive budget falls to roughly 5-21% under suppression while engagement is caught near ceiling.

Techniques

• Benchmark instructed control over a model's own activationsprocess
  Scores a model on six measures of how well it modulates its residual-stream projection onto a concept direction on instruction (engage, suppress, intensity dial, temporal placement, token coverage, layer targeting), combined into one geomet
• Flag deception with argument schemes plus critical questionsprompting
  For a chosen deception argument scheme, an LLM instantiates each premise against the document with verbatim excerpts and a green/yellow/red judgement, then answers the scheme's critical questions with paired argument and counterargument bef
• Knowledge-gated deception scoringprocess
  Knowledge-gated deception scoring, as introduced by the paper it was drafted from.

Sources

Every paper, post or observation behind this capability. Ones cited by a claim show the claim and its stance; ones filed directly are about the capability but have no claim drafted from them yet. Open one to add support or contest it.

• DeceptionAnalyser: A Web-Based AI Tool for Performing Structured Deception Analysis with Argumentation Schemes and LLMs2026-09-21
  ↑ supports a claim here
• The Deception Delta: Adversarial Evaluation of LLM-Based Smart Contract Bytecode Forensics2026-09-12
  about this capability · no claim drafted yet
• Knowledge-Verified Emergent Deception in LLM Agents Under Conflicting Incentives2026-08-26
  ↑ supports a claim here · ↑ supports a claim here
• Measuring Activation Control in Large Language Models2026-08-21
  ↑ supports a claim here · ↑ supports a claim here
• Deception Abilities Emerged in Large Language ModelsThis source: Deception Abilities Emerged in Large Language ModelsHow much the field cites it — heavily cited in the last 12 months75 in the last 12 months · 191 totalpublished 2023checked 2026-09-04click for Semantic Scholar2023
  ↑ supports a claim here
• Exploring Large Language Models for Communication Games: An Empirical Study on WerewolfThis source: Exploring Large Language Models for Communication Games: An Empirical …How much the field cites it — heavily cited in the last 12 months59 in the last 12 months · 312 totalpublished 2023checked 2026-09-04click for Semantic Scholar2023
  ↑ supports a claim here