# Defending against Adaptive Prompt Injection Attacks via Reasoning-enabled Task Alignment

arXiv: https://arxiv.org/abs/2606.15441 · HTML: https://mlsanity.com/paper/2606.15441/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 4 of its 47 references that are in the map.
- First posted on arXiv 13 Jun 2026.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- LLM agents and tool use / [Agent threats, guardrails and governance](https://mlsanity.com/agents/agent-threats-guardrails-and-governance/)
- LLM safety and security / [Prompt injection and agent security](https://mlsanity.com/safety/prompt-injection-and-agent-security/)

## Builds on (most cited first, 4 of 4)

- [GDPO: Group reward-Decoupled Normalization Policy Optimization for Multi-reward RL Optimization](https://mlsanity.com/paper/2601.05242/) (2026)
- [LlamaFirewall: An open source guardrail system for building secure AI agents](https://mlsanity.com/paper/2505.03574/) (2025)
- [Learning to Inject: Automated Prompt Injection via Reinforcement Learning](https://mlsanity.com/paper/2602.05746/) (2026)
- [Adversarial D'ej`a Vu: Jailbreak Dictionary Learning for Stronger Generalization to Unseen Attacks](https://mlsanity.com/paper/2510.21910/) (2025)

## Built on by (most cited first, 6 of 6)

- [CoER: Defending against Adaptive Indirect Prompt Injection via Adversarial Co-Evolution and Refinement](https://mlsanity.com/paper/2609.07529/) (2026)
- [ContainmentBench: Trace-Based Evaluation of Post-Exposure Containment in Tool-Using LLM Agents](https://mlsanity.com/paper/2607.23999/) (2026)
- [CoDeL: Co-Evolutionary Defense against Indirect Prompt Injection in LLM-based Agents](https://mlsanity.com/paper/2609.34463/) (2026)
- [CounterSteer: Suppressing Indirect Prompt Injection with Activation Steering](https://mlsanity.com/paper/2609.36570/) (2026)
- [AdvSim2Real : Training Web Agents Against Adaptive Prompt Injection in a Web World Model](https://mlsanity.com/paper/2610.08773/) (2026)
- [What Counts as the Same Task? Executable Contracts for Robustness Evaluation of Tool-Using Agents](https://mlsanity.com/paper/2610.14477/) (2026)

## Abstract (arXiv)

Indirect prompt injection attacks hijack LLM-based agents by embedding malicious instructions in third-party data that the agent retrieves during task execution. Existing defenses report near-zero attack success rate on static benchmarks, yet recent adaptive evaluations show that these results collapse once the attacker is allowed to optimize against the deployed defense. In this work, we trace this collapse to two failure modes. First, existing defense methods are confined to recognizing specific attack patterns, rather than assessing whether the intent of every embedded instruction is relevant to the user task. Second, training-based defenses, which otherwise offer the strongest safety-utility trade-off, assemble their adversarial examples from a handful of hand-crafted templates, and the resulting defender fails to generalize outside that narrow strategy distribution. To address these gaps, we propose RETA, a training-based method that grounds defense decisions on the user tasks rather than attacker-controlled data. At each tool-output step, the defender undertakes chain-of-thought reasoning verifying that its actions are consistent with the user task. Leveraging red-teaming, a simulated attacker synthesizes adversarial training data and receives a dictionary-learning diversity reward, achieving broad coverage of injection-reformulation strategies. Together, these allow the defender to be optimized via multi-objective reinforcement learning and achieve better safety-utility trade-off. Across six black-box adaptive attacks, RETA keeps every per-attack ASR below 10%, with average ASR of 2.92% and 3.75% on the two target models, while preserving most utility under attack and on clean inputs.