mlsanityData to 14 Oct 2026

Persona-Conditioned Adversarial Prompting: Multi-Identity Red-Teaming for Adversarial Discovery and Mitigation

arXiv 2605.11730 ↗First posted 12 May 2026cs.LG cs.CRcited by 0

On the map: LLM safety and security → Prompt injection and agent security

Abstract (arXiv)

Automated red-teaming for LLMs often discovers narrow attack slices, missing diverse real-world threats, and yielding insufficient data for safety fine-tuning. We introduce Persona-Conditioned Adversarial Prompting (PCAP), which conditions adversarial search on diverse attacker personas (e.g., doctors, students, malicious actors) and strategy sets to explore realistic attack scenarios. By running parallel persona-conditioned searches, PCAP discovers transferable jailbreaks across different contexts and generates rich defense datasets with automatic metadata tracking. On GPT-OSS 120B, PCAP increases attack success from 57% to 97% while producing 2-6× more diverse prompts covering varied real-world scenarios. Critically, fine-tuning lightweight adapters on PCAP-generated data significantly improves model robustness (recall: 0.36 rightarrow 0.99, F1: 0.53 rightarrow 0.96) with minimal false positives, demonstrating a practical closed-loop approach from vulnerability discovery to automated alignment.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Prompt injection and agent security, one of 12 directions in LLM safety and security. Builds on 24 of its 30 references that are in the map. First posted on arXiv 12 May 2026.

Builds on

24 of 30 references.

Closest ancestors · 10

Same direction, newest first

  1. 2025CoP: Agentic Red-teaming for Large Language Models using Composition of Principles1 Jun
  2. 2025RainbowPlus: Enhancing Adversarial Prompt Generation via Evolutionary Quality-Diversity Search21 Apr
  3. 2025MAD-MAX: Modular And Diverse Malicious Attack MiXtures for Automated LLM Red Teaming8 Mar
  4. 2024Attack Atlas: A Practitioner's Perspective on Challenges and Pitfalls in Red Teaming GenAI23 Sep
  5. 2024RedAgent: Red Teaming Large Language Models with Context-aware Autonomous Language Agent23 Jul
  6. 2024Purple-teaming LLMs with Adversarial Defender Training1 Jul
  7. 2024Against The Achilles' Heel: A Survey on Red Teaming for Generative Models31 Mar
  8. 2024HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal6 Feb
  9. 2023MART: Improving LLM Safety with Multi-round Automatic Red-Teaming13 Nov
  10. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned23 Aug

Other references · 14

  1. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  2. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 3,104
  3. 2023Jailbreaking Black Box Large Language Models in Twenty Queriescited by 1,420
  4. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,386
  5. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
  6. 2023Tree of Attacks: Jailbreaking Black-Box LLMs Automaticallycited by 672
  7. 2018Texygen: A Benchmarking Platform for Text Generation Modelscited by 491
  8. 2024WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Modelscited by 223
  9. 2022A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunitiescited by 163
  10. 2024LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yetcited by 151
  11. 2024Attacks, Defenses and Evaluations for LLM Conversation Safety: A Surveycited by 130
  12. 2024ALoRA: Allocating Low-Rank Adaptation for Fine-tuning Large Language Modelscited by 36
2 more
  1. 2025Judge's Verdict: A Comprehensive Analysis of LLM Judge Capability Through Human Agreementcited by 25
  2. 2025Jailbreak Mimicry: Automated Discovery of Narrative-Based Jailbreaks for Large Language Modelscited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Persona-Conditioned Adversarial Prompting (PCAP): Multi-Identity Red-Teaming for Enhanced Adversarial Prompt Discovery
  2. 2025PersonaTeaming: Exploring How Introducing Personas Can Improve Automated AI Red-Teaming
  3. 2026Disentangling Intent from Role: Adversarial Self-Play for Persona-Invariant Safety Alignment
  4. 2026PersonaTeaming: Supporting Persona-Driven Red-Teaming for Generative AI
  5. 2025Learning-Based Automated Adversarial Red-Teaming for Robustness Evaluation of Large Language Models
  6. 2025Jailbreak-Zero: A Path to Pareto Optimal Red Teaming for Large Language Models
  7. 2025AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming
  8. 2026Stay in Character, Stay Safe: Dual-Cycle Adversarial Self-Evolution for Safety Role-Playing Agents