The main problem with the ChatGPT AI Agent and similar solutions is that services are already starting to defend against them, as expected. Users are already complaining that the new agent from OpenAI initially logged in, filled out spreadsheets, etc., but now it specifically encounters pop-ups and other elements that confuse it, causing it to simply stop performing tasks. It's understandable—why would Google with their Gemini need a competitor?
In the end, I think this will only be resolved through legal disputes that will take years, while everyone will build even bigger walls around themselves under the pretext of caring for users, of course.
It's a shame.
