Beyond Anthropic: OpenAI and Google Also Embed Watermarks in LLM Outputs
StewartalsopIII · x · 2026-08-11
A user pointed out that beyond Anthropic, both Google and OpenAI have signed agreements to embed watermarks in the outputs of their closed-source LLMs. The poster warned against trusting any closed-source models, suggesting that providers could secretly embed spyware or tracking mechanisms as they see fit.
More from Safety
- Elite security team benchmarks 8 AI agent sandboxes, exposing escape risks — ycombinator · 2026-08-11
- OpenAI Launches Cyber-Trained AI Model Amid Rising AI-Led Attacks — TechCrunch AI · 2026-08-11
- The AI Safety Paradox: Adversarial Training Data as a Trojan Horse — alexbilz · 2026-08-11
- 80,000 Hours Deep Dive: Why Power-Seeking AI Poses an Existential Risk — AndyMasley · 2026-08-11
- Research Explores Re-identification and Surveillance Risks Posed by Agentic LLMs — QVeraLiao · 2026-08-11
- EU Unveils Official Icons for Labeling AI-Generated Content Under AI Act — Polymarket · 2026-08-11