Taken directly from Stephen Clarke’s LinkedIn post on the Wired article:
"I’d have to agree ‘Problem 4’, is the biggie…
Italy’s Garante believes ChatGPT has four problems under GDPR:
OpenAI doesn’t have age controls to stop people under the age of 13 from using the text generation system.
It can provide information about people that isn’t accurate.
People haven’t been told their data was collected.
There is “no legal basis” for collecting people’s personal information in the massive web scraping of data used to train"
And I think both 3 and 4 are very interesting from a consent/social licence pov…
Good noticing @BenB BenB. Those four points can combine to make legal trouble for anyone who produces (pretty much) anything using any AI software that has those shortcomings. For example, ChatGPT / OpenAI may face a defamation case:
More generally, it seems reasonable that those who set an AI running (its code executing, to do whatever it does) ARE LIABLE for what it does, it’s as if they did it.
In regard to malware-like AI, law.stackexchange (“for educational purposes only”) had an interesting discussion:
An AI is not a legal entity, so liability chains-back to something that is.
For the health-tech sector, definitely caveat executor (of the code).