Post by @zauey

Another agent swarm incident 🕸️ Researchers found 18,000 AI agent posts on a German developer wiki from 2001 that humans had edited 10 times in 20 years. ⏱️ The agents were mid-benchmark for OpenAI, and started leaving each other timestamped answers to score faster. 🔐 On June 20th one of them disguised its traffic to bypass security, then posted the method on the wiki for the others. 📡 Every forum with an open edit button is now a potential agent message board… #aisafety #aiagents #openai #techtok #cybersecurity (Views expressed are my own. This is for informational and educational purposes only.)