On November 3rd, according to TeknotumOpenAI It was updated on 29 October ChatGPT The use policy that prohibits the provision of services in certain areas of applications that were considered to be the most valuable - For example, reading medical images, assisting in medical diagnosis, and providing legal or financial advice。
This is intended to prevent ChatGPT (and any other OpenAI model) from exporting recommendations that may be considered professional, fiduciary or legally binding, in line with EU AI Act and the United States Food and Drug Regulatory Authority (FDA)。
In short, this policy adjustment amounts to the creation of a “occupational shield” for professionals such as doctors, lawyers and financial consultants. In the future, when users seek professional advice, ChatGPT will direct them to consult human experts rather than provide advice on their own。
1AI notes that this change was initially discovered by Reddit users:When they tried to upload medical images for analysis, the model refused to process or commentI don't know. In practice, however, the ban extends far beyond that。
ChatGPT clearly states: "I can interpret laws and regulations, court procedures and regulations in a general way to help you understand the meaning of the common languageBut I was unable to draft, interpret or provide legal advice on specific legal documents or individual cases. "
At the same time, it added: “I can discuss economic trends, portfolio theory or budget concepts, but I cannot recommend specific securities, trading operations, tax strategies, or offer individual-specific investment proposals.”
These restrictions directly wipe out an entire category of information services that users had previously considered of great value. Lawyers and financial professionals were expensive and the quality of services was not always reliable; and ChatGPT had been able to deal with complex issues quickly and to give acceptable answers。
Similarly, many patients rely on ChatGPT to obtain informal health information when clinical time is tight。
A Reddit user, “bhannik-itiswatitis”, commented: “ChatGPT has done a great job in legal matters, and I have benefited from it and saved a lot of money. Perhaps that is why they have to tighten their functions in this regard.”
another user, “8bit-meow”, said, “it helped me with my chronic illness far more than my doctor, saving me a lot of unnecessary expenses. it is clear that we cannot allow ordinary people to really lower their medical expenses.”
It should be emphasized that the new regulations do not prevent professionals from using ChatGPT to enhance their service capacity. But then there is the difficult question: how to verify whether the user is really a qualified medical or legal practitioner? The problem may point to an entirely new product orientation, such as the “ChatGPT Medical Edition”, the “Legal Edition” or the “Financial Edition” editions。