While the slandering of Hood's reputation is troublesome, it will be very difficult to prove an AI algorithm is at fault for disseminating defamatory information. To defame someone — legally speaking — the perpetrator must have knowingly disseminated the falsities with malice, but how could a computer do that? Such cases involving public figures will lead to ever-growing debate on the issue of AI and its role in public discourse, but to sue a robot isn't a winnable course of action.
The case against OpenAI has nothing to do with the algorithm and everything to do with the company's delayed response to Hood's request. Once Hood proved the information to be false, OpenAI should have scraped it off the platform immediately, but, according to the lawyers, it hasn't done that and therefore has opened itself up to a legitimate allegation of defamation.
There's a 50% chance that OpenAI will announce GPT-5 by April 2024, according to the Metaculus prediction community.