phenomenon · first seen 9 Oct 2026
prompt inference attack
Also: prompt leakage
Attacks that recover a user's private prompts or inputs from observed model execution.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Physical Side-Channel Attacks on Edge LLMs: Extracting Output Tokens and User PromptsPapers · 9 Oct 2026 · Najeh Alhalawani, Gunes Acar, Zhuoran Liu, Lejla Batina
Latest sources
- Physical Side-Channel Attacks on Edge LLMs: Extracting Output Tokens and User PromptsPapers · 9 Oct 2026 · Najeh Alhalawani, Gunes Acar, Zhuoran Liu, Lejla Batina