I wrote the AI policy for a $2B company which makes networking products (I’m head of Risk Management and Cybersecurity) . Many technologies can be used for good or bad - eg nuclear energy, AI, guns, drugs, etc. The pitfalls, levels of reliability, and effective use are not really that difficult to figure out.

What is difficult is trusting the people who are the stewards of this technology. It comes down to their intentions. Most people in management are interested in power vs doing what is in the best interests of a broad group of people.

One of the most important principles in risk management and cyber security is transparency. But that can only be achieved if it’s done in an environment where Mangement cares about transparency and actually wants it. So ultimately this all comes down to the senior Mangement and culture of the company. Or the government.

That will drive things much more than anything else.