Heads up sora will get you banned

OpenAI has begun enforcing its terms of service against users who generated content with its Sora video model, leading to unexpected and retroactive account bans. According to user reports on social platforms like Reddit, long-term ChatGPT Plus subscribers are finding their accounts terminated for Sora videos created during the model's initial release window, sometimes months prior. One user stated they were banned for "dumb Rick and Morty videos (nothing sexual)" made "the week it came out," despite being a paying customer "since the option to pay existed." This suggests OpenAI is applying its content policies retrospectively to early Sora outputs, catching users off guard.

The enforcement wave has sparked significant user backlash and migration. Affected professionals, who relied on ChatGPT for work, are expressing frustration over the lack of warning and the perceived inconsistency in policy application. The incident has practical consequences, as one banned user noted they've switched to "Gemini and its far better," highlighting a competitive risk for OpenAI. This situation underscores the complex challenges AI platforms face in moderating synthetic media after the fact and the potential erosion of user trust when enforcement actions feel sudden and disproportionate to the perceived violation.