User warns OpenAI's 5.6 Sol model flags harmless tasks, causing bans

A Reddit user reports being banned after using OpenAI's 5.6 Sol model for a legitimate Excel task, which was flagged as a cybersecurity threat. The user's appeal was rejected within 2 hours, raising concerns about overly aggressive safety filters in the new model.