OpenAI unveiled its GPT-5.5 model on Monday, achieving the state-of-the-art across 14 benchmarks including economic knowledge and specialized cybersecurity. The model narrowly beat Anthropic’s Claude Mythos on Terminal-Bench 2.0, reclaiming the lead for OpenAI in the race for agentic intelligence. Alongside this release, the company deployed 'ChatGPT for Clinicians,' a specialized tool offered for free to American doctors to automate diagnostic work. According to researchers, this move bypasses hospital IT systems and creates a structural dependency on private algorithms for medical care. While the model is offered as a public service, critics describe it as a 'Cognitive Enclosure' of the medical commons, where the expertise of human practitioners is harvested and then replaced by automated agents. The FDA is currently accelerating the review of three psychedelic drugs for depression, but the underlying infrastructure of care is being handed to tech giants. As clinicians increasingly rely on the free tier of these models, the traditional practice of medicine is being deskilled and gated behind high-priced corporate subscriptions. The free software is not a gift; it is a vacuum designed to suck the remaining human agency out of the clinic and into the server farm.