After being heavily criticized by privacy experts and pundits, ChatGPT has finally released a feature in which users can take advantage of the tool while opting out of sharing their input. This has quite a few implications:
First, from the perspective of internet privacy advocates, this is a win. However, it does not remove other privacy concerns regarding chatbot AI. Even though you could now opt out of sharing your data before you used the product, this does not mean that you could retroactively delete your data after you have already shared it with the service, and especially not after it has been used to train the AI. I think this will inevitably continue to raise a “right to be forgotten” issue with respect to these services.
Second, for workplaces already using AI tools, such a function to merely take advantage of the algorithm might be a boost in efficiency. For example, the user will no longer need to anonymize the instructions (e.g. “write an email to John Doe to submit my final essay for Media and Entertainment law”) when they do not want to share specific information. Of course, this will still depend on how sensitive the information is as users should still be concerned about the risk of data breaches.
Third, it may be a loss for those trying to police and prevent people from taking advantage of generative AI in creative processes. For example, universities may have even more difficulty finding out whether a work submitted by a student was written using AI tools.