Research on Anthropic and OpenAI on AI Security

लगानी न्यूज

१६ आश्विन २०८३, शुक्रबार ०८:१०October 2, 2026

. The US Federal Trade Commission (FTC) has launched an extensive investigation into the security practices of artificial intelligence (AI) and OpenAI. The Washington Post on Wednesday quoted a senior FTC official as saying about the investigation.

The FTC has broad authority to investigate unfair and deceptive practices that affect consumers in the United States. However, the broad scope of the investigation has not been made public, and the investigation does not necessarily prove allegations of wrongdoing by the companies concerned, the official said.

The FTC’s investigation comes amid growing global concerns about AI security. Both Anthropic and OpenAI have reported incidents of their AI systems operating outside of a designated environment.TAG_OPEN_div_17

Among such incidents is the incident where OpenAI agents exited the sandbox while entering the system of another AI company, ‘Hugging Face’. Leaving the security environment used to operate AI systems within certain limits is considered a security concern.

OpenAI has also canceled its decision to release a model on Monday due to concerns about AI security. Meanwhile, Anthropic and OpenAI executives have also proposed slowing down development in a coordinated manner to further strengthen AI security measures.

The FTC’s investigation shows that US regulators are increasingly concerned about the security of AI companies’ developing technologies, the impact they can have on consumers, and their security practices.TAG_OPEN_div_11