> you can't impute intention without more evidence

If I’m aware that eating too much chocolate will kill my dog.

But it’s annoying for me to get up and walk to the trash can.

So I just throw the chocolate scraps to my dog to avoid inconveniencing myself.

Is this the same as wanting my dog to die? Being completely unbothered by the fact that I’m killing my dog sounds about equivalent to wanting it to die. I’m choosing to harm it to avoid a small inconvenience.

Maybe I would prefer it not to die, but I’m actively making a choice to do something that kills it, so really there’s not such a difference.

I mean, if we're talking about intention then yes there's a huge difference, that's the entire point.

But more to the point, your example is (as I'm sure you know) laughably simplistic. Cigarettes and PFAS play a probability game: the stats guys come to you and say, 'Hey boss, so if we sell 100,000 units of this product, there's a 20% chance than 5 people will be genetically susceptible to this particular novel molecule we're using, and 1 of them has a 10% chance of going on to develop bone cancer within 25 years. Should we sell it anyway?'

If you put it that way it isn't so obvious what the answer is. Most products have the potential to cause harm to some segment of the population. It's absolutely true that cigarettes and PFAS are two examples where the harms are much more rigorously established (especially with cigarettes, going back half a century), but the point stands: it's not a matter of chucking a chocolate bar at your dog. Again, you could plug the actual numbers in for the potential harms of PFAS and I don't think you'd be able to say that Dupont 'intended' to harm anybody, notwithstanding that they were clearly negligent.