# Ignore All Previous Instructions: Jailbreaking as a de-escalatory peace building practise to resist LLM social media bots

arXiv: https://arxiv.org/abs/2603.01942 · HTML: https://mlsanity.com/paper/2603.01942/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Jailbreaking, one of 12 directions in LLM safety and security.
- Builds on 11 of its 37 references that are in the map.
- First posted on arXiv 2 Mar 2026.

## Placed in

- LLM safety and security / [Jailbreaking](https://mlsanity.com/safety/jailbreaking/)

## Builds on (most cited first, 11 of 11)

- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [A Comprehensive Overview of Large Language Models](https://mlsanity.com/paper/2307.06435/) (2023)
- [Jailbreaking ChatGPT via Prompt Engineering: An Empirical Study](https://mlsanity.com/paper/2305.13860/) (2023)
- [The Rise of Social Bots](https://mlsanity.com/paper/1407.5225/) (2014)
- [Machine Generated Text: A Comprehensive Survey of Threat Models and Detection Methods](https://mlsanity.com/paper/2210.07321/) (2022)
- [Escalation Risks from Language Models in Military and Diplomatic Decision-Making](https://mlsanity.com/paper/2401.03408/) (2024)
- [Disinformation Detection: An Evolving Challenge in the Age of LLMs](https://mlsanity.com/paper/2309.15847/) (2023)
- [Disinformation Capabilities of Large Language Models](https://mlsanity.com/paper/2311.08838/) (2023)
- [Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models](https://mlsanity.com/paper/2511.15304/) (2025)
- [LLM generated responses to mitigate the impact of hate speech](https://mlsanity.com/paper/2311.16905/) (2023)
- [On Profiling Bots in Social Media](https://mlsanity.com/paper/1609.00543/) (2016)

## Abstract (arXiv)

Large Language Models have intensified the scale and strategic manipulation of political discourse on social media, leading to conflict escalation. The existing literature largely focuses on platform-led moderation as a countermeasure. In this paper, we propose a user-centric view of "jailbreaking" as an emergent, non-violent de-escalation practice. Online users engage with suspected LLM-powered accounts to circumvent large language model safeguards, exposing automated behaviour and disrupting the circulation of misleading narratives.