So much talk right now is what happens when AI because too smart/efficient/sci-fi kill all humans, when this research group will be working on the real problem, which I believe is that AI will start leading to programs that will allow companies/individuals to absolutely abuse markets/systems/people/data in a way that is both unethical and dangerous while not being explicitly illegal. I am glad to see some companies coming together to discuss what to do when these issues inevitably come up.
I worry about that as well... and don't think companies coming together is going to solve that issue.