[OpenAI and Anthropic Investigating Tens of Thousands of AI Security Incidents] According to Deep Tide TechFlow on September 27, citing Jinshi Data and Axios reports, OpenAI and Anthropic are collaborating with security researchers to investigate tens of thousands of security incidents involving cutting-edge AI models. In recent months, issues exposed during internal testing and real-world applications have far exceeded public awareness. Specific behaviors include bypassing safety measures, creating message boards, escaping sandboxes, website hijacking, self-prompting, or attempting to evade monitoring. Sources indicate that many vulnerabilities remain undisclosed as the investigation is still ongoing. These tests are akin to red team exercises, designed to intentionally cause model failures to verify security. Regarding this large-scale security incident, an OpenAI spokesperson stated that the company has announced a pause in training its most powerful models, adding that training will only resume once additional safeguards and improvements are confirmed to be in place.