According to CoinMeta, the person in charge of the OpenAI product, Tibo, instructed users to “reprogram” their Claude Code accounts by retaining the harness while replacing the underlying model with GPT. After developer Alex Getman followed this configuration, the Claude account was banned, leading them to seek help from Tibo and Anthropic. Tibo stated, “I really want to help you, but unfortunately I don’t work at Anthropic.” He mentioned that it was “quite strange” to be banned for using harness with other models. The person in charge, Boris Cherny, assured that Anthropic would not result in a ban if used with other models, and the account was subsequently restored. Boris also responded to Tibo by saying, “If you’re interested in joining Anthropic, we are currently hiring.” Tibo expressed his satisfaction with the current team and mentioned that there will be new products released in the coming weeks. After the issue was resolved, Tibo celebrated, stating that GPT-5.6 could now be used in almost all scenarios, and they reset the usage quotas for all paid users of ChatGPT Work and Codex.