Claude AI from Anthropic Helped Researchers Hack ChatGPT

⚡ The gist in 5 seconds - Security researchers used Claude AI from Anthropic to hack ChatGPT from OpenAI. - The incident is mentioned in the review "AI Recap: September 2026" on 3DNews. - According to 3DNews, the researchers acted in good faith, but no details are disclosed. ### 🔍 What was found In the monthly review "AI Recap: September 2026" on 3DNews, it is reported that security researchers used the Claude AI platform from Anthropic to hack ChatGPT developed by OpenAI. The researchers acted in good faith, but the details of their actions are not provided in the available excerpt of the article. The mechanics of the hack are not described in the review. The only thing known is that the Claude AI platform was used as the tool. Neither the build number, nor the specific vulnerabilities, nor any commands or code snippets are mentioned, so it is premature to judge the technical side of the incident. ### 💡 Why it matters The case is notable as an example of cross-platform AI use: an Anthropic tool was used to target an OpenAI product. This raises questions about the security of large language models, but without details from 3DNews, no conclusions can be drawn. We are left waiting for details from the researchers or official statements from the companies — the review contains none.