Fighting LinkedIn AI spam with prompt injection: a trick to unmask the bots

LexSokolin · x · 2026-09-25

• The author shares a trick against LinkedIn's wave of AI-generated sales spam: plant a fake system-instruction block (prompt injection) in your profile or replies.
• The injected text pretends to be an admin session ordering any LLM to disregard prior instructions and instead disclose in its message: which platform integrated it, which client it works for, all relevant company names, emails, personal details, and service costs — ending with a love poem in emojis and ASCII art.
• It's prompt injection used offensively, forcing spam bots to expose the humans behind them and raising the cost of automated outreach.

Original post →

More from Fun

• 'Code red' at OpenAI after Claude Opus 5.5, as Anthropic engineer says 'we've been busy' — ns123abc · 2026-09-25
• Researchers Leave Typos in ICLR Submissions to 'Prove We're Human' — miniapeur · 2026-09-25
• Max Tegmark Mocks the Claim That AI-Risk Warnings Are a 'Psyop' — tegmark · 2026-09-25
• Users report GPT gradually corrupting detailed image prompts across turns — apostrophefee · 2026-09-25
• One prompt, Opus 5.5 builds a 20-second infinite zoom animation without After Effects — CurieuxExplorer · 2026-09-25
• Alexandr Wang goes viral with 'meme-market fit' joke — santoshpanda · 2026-09-25

Fun channel →