Can an "AI ethics body" set up by companies that develop and use AI software possibly be adequate?

I have no solid proof, but I doubt it.