The major Ai providers (frontier models) have contract clauses and marketing copy that specifically say they do not use your data or sessions for their training.

Microslop / GitHub is a notable deviation here.

▲

JohnFen 7 hours ago | parent [-]

True, but I'd be very hesitant to trust such statements/contract clauses.

▲

verdverm 5 hours ago | parent [-]

Why would a company that is not in the same line of business risk the lawsuits and brand damage?

This belief you hold doesn't make sense from any logical perspective

	▲	JohnFen 3 hours ago \| parent [-]
		The lawsuit risk isn't huge. First, most people probably wouldn't be able to discover this has been done. Second, most people wouldn't be able to afford to sue, and the company can just settle with those who can. I'm also not sure that the risk of brand damage is that compelling. A whole lot of companies misbehave and don't suffer enough to stop their behavior because of it. genAI companies particularly. Overall, the things they have already said and done strongly indicate to me that they are not trustworthy. That's the main source of my doubt. Why would I expect them to be any different tomorrow than they were yesterday?