Strategic Deception under Information Asymmetry

15 concepts7 papers1 proposal

In classical game theory, we often assume players have a clear view of the rules, the environment, and their opponents' goals. However, in real-world interactions—ranging from military engagements to automated negotiations—information is rarely distributed evenly. This "information asymmetry" is not merely a hurdle to overcome; it is a potent strategic lever. When one agent knows more than another, they can move beyond simple optimization to strategic deception: the intentional manipulation of an opponent’s beliefs to steer them toward suboptimal or even hazardous choices.

The Mechanism of Misperception

The core intuition behind strategic deception is that an agent’s actions are dictated by their perception of the environment, not the environment itself. If an informed agent can influence that perception, they can effectively control the opponent's behavior.

In reachability games, for instance, Kulkarni et al. demonstrate that an agent can exploit an opponent's misperception of their capabilities. By appearing weaker or more restricted than they truly are, the informed agent can lure the opponent into a false sense of security. The authors show that this tactic can transform "losing" states—where the opponent would normally win under perfect information—into "winning" states for the deceiver.

This principle extends to physical environments. In differential games, such as pursuit-evasion scenarios, Comandur et al. explore how a more-informed player can maximize a "deception payoff." By manipulating the opponent's understanding of the environmental dynamics, the deceiver can lead a less-informed agent into dangerous situations, such as collisions, which the victim would have easily avoided had they possessed accurate information.

Steering Negotiations through Epistemic Advantage

Negotiation is perhaps the most common arena for this exploitation. When agents bargain over resources or goals, the "fairness" of an outcome is often subjective and dependent on knowing what the other party values.

De Clercq et al. highlight that in Boolean games, agents with superior knowledge of their counterparts' goals consistently secure better outcomes. Because the less-informed agents cannot verify if a proposed agreement is truly a fair compromise, they are often forced to accept suboptimal deals. Similarly, Erlich et al. show that even simple knowledge of an opponent’s ordinal preferences (their ranking of outcomes) allows an informed party to steer the final agreement toward their own ideal result, far outperforming standard "maxmin" strategies that assume no such knowledge.

The Fragility of Cooperation and Learning

While communication is often viewed as a tool for fostering cooperation, information asymmetry can turn it into a vector for exploitation. In public goods games, where agents must decide how much to contribute to a shared pool, Orzan et al. found that informed agents leverage communication to deceive less-informed peers about game rules or incentives. This deceptive signaling increases the rewards for the deceiver but undermines the collective benefit.

This creates a fundamental challenge for multi-agent learning. For learning to be stable, agents usually need to reach an equilibrium. However, Brafman et al. prove that in environments with "imperfect monitoring"—where agents cannot see each other's payoffs—efficient learning equilibria are impossible unless the agents share common interests. The asymmetry allows for "strategic lying" about rewards, preventing the system from ever settling into a stable, fair state.

Incentives for Corruption

Finally, this asymmetry changes the calculus of security. In two-party computation protocols, Wang et al. argue that attackers are not just "malicious" but rational. An adversary with superior information about the reliability of participants can more accurately predict the success of a corruption attempt. The information advantage itself creates a positive incentive for the attack, as the adversary can exploit the participants' mutual uncertainty to maximize the profitability of the breach.

Across these domains, the research suggests a unified theme: information is not just data—it is the foundation of the "game" being played. By controlling that foundation, a strategic agent can rewrite the rules of the interaction to their own advantage.

Key Foundational Work

Multilateral Negotiation in Boolean Games with Incomplete Information Using Generalized Possibilistic Logic (2015) — contributed 2 concept elements to this direction.
Efficient Learning Equilibrium (2002) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_exploiting_opponent_misperception_e1d538ec,
  title = {exploiting opponent misperception},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/exploiting-opponent-misperception/rdr_e1d538ecd64247679a78eb63d927db61},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_exploiting_opponent_misperception_e1d538ec_v1,
  title = {exploiting opponent misperception},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/exploiting-opponent-misperception/rdr_e1d538ecd64247679a78eb63d927db61/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake