Indeed, points out Wendell Wallach, a senior fellow at the Carnegie Council on Ethics in International Affairs, where he is one of the directors of this organization. AI and Equality Initiativeand author of The Dangerous Craftsman: How to Keep Technology from Slipping Out of Our Control, some industries already have some regulations in place that should also govern artificial intelligence.
When it comes to management performance, regulations already exist – in some industries more than others. “Health care is already pretty well regulated, at least in the United States and Europe,” Wallach says. “In a sense, many AI programs will be governed by what already exists. So the question is what additional ones [regulations] Do you need? You probably need different forms of testing and compliance in some areas.
For example, US health care regulations state that a doctor has a duty of care; if they misdiagnose the patient, they may be liable for medical malpractice. If a doctor relies on AI to help diagnose patients and misdiagnoses them, they could still be held liable (although experts note that there is a likely rebuttal question: will it be common to blame the AI for the error). But that means managing the result – a misdiagnosis.
And many experts believe that even this does not go far enough. More safeguards need to be put in place to regulate the processes themselves: for example, quality control assessments of key data that led to misdiagnosis.
“There will have to be different ways to demonstrate that AI is implemented responsibly. And the question is, how do you implement the tests? How do you compare them?” Singapore's Communications and Information Minister Josephine Teo said in Davos panel. “These things are still very much in the making. No one has the answers yet.”
Overall, experts say, the main goal is to ensure that AI is used well, regardless of the approach.
“If you think about the work ahead of us – to tackle the climate crisis, to improve people's health and well-being, to ensure that our children are educated and that people at work can learn new skills… look at how we will do it without the power of AI,” – said Prabhakar. “From an American perspective, we've always thought of doing this regulatory work as a means to that end — not just to protect rights, which is absolutely necessary not only to protect national security, but also to accomplish these big goals.”
“Bacon fanatic. Social media enthusiast. Music practitioner. Internet scholar. Incurable travel advocate. Wannabe web junkie. Coffeeaholic. Alcohol fanatic.”