Dev ships Simple Unmark in 2 days: a tool to strip AI text watermarks
haltakov · x · 2026-08-20
Built in under two days, Simple Unmark removes watermarks from AI-generated text. It handles two mechanisms: statistical watermarks from Gemini/Claude-style token-choice patterns (SynthID-like), addressed by rewriting syntax, clause order, transitions and vocabulary; and hidden Unicode copy-paste artifacts, removed exactly. The site offers free local scanning as you type, an invisible-character inspector, and vendor-specific pages (Gemini, Claude, SynthID) that explain each mechanism and admit no public detector exists to verify results.
More from Safety
- Experts criticize OpenAI safety strategy, call for independent oversight — andersonbcdefg · 2026-08-20
- Criticism: AI detectors flag old writing and everything as biothreats — bratton · 2026-08-20
- Varonis Bets on Auto-Remediation to Compete with Free Microsoft Purview — shashib · 2026-08-20
- MATS Program Launches New Stream on AI Auditing, Eval Standards, and Governance — Miles_Brundage · 2026-08-20
- Hacks cost just $2 in DeepSeek tokens and an AI agent — Thionne_WTZ · 2026-08-20
- RAG poisoning creates false confidence: monitoring attention beats uncertainty checks — rohanpaul_ai · 2026-08-20