7 Key Principles for Effective AI Regulation

Seven Principles for Responsible Regulation

Companies in democracies have been leading the charge in AI, but we can't rest on our laurels. We need to keep pushing for future AI breakthroughs because, let's face it, we're ahead in some areas but lagging in others.

To keep the innovation train rolling while keeping things responsible, we suggest these seven principles for AI regulation:

1. Support responsible innovation. The Senate's roadmap kicks off by calling for more cash for both AI innovation and safety measures. That's smart because they go hand in hand. Better tech means safer systems, and while there's always some uncertainty with new tech, we can still build trust without slowing down the good stuff.
2. Focus on outputs. Let's push for AI that churns out quality stuff while keeping the bad stuff at bay. By zeroing in on what AI actually produces, regulators can tackle issues head-on without getting bogged down in the nitty-gritty of computer science. That keeps things grounded in real-world problems and avoids overdoing it with regulations that might stifle AI's potential.
3. Strike a sound copyright balance. Fair use and copyright exceptions are great for science and learning, but website owners should be able to opt out of having their content used for AI training with machine-readable tools.
4. Plug gaps in existing laws. If something's illegal without AI, it's illegal with AI. No need to reinvent the wheel; we just need to fill in the gaps where current laws don't quite fit AI.
5. Empower existing agencies. There's no one-size-fits-all for AI regulation, just like there's no single law for all uses of electricity. We need to beef up existing agencies and make sure they're all AI-savvy.
6. Adopt a hub-and-spoke model. This model sets up a central hub of tech know-how at a place like NIST to help the government get a better handle on AI and support different sectors, recognizing that banking issues are different from those in pharmaceuticals or transportation.
7. Strive for alignment. With so many AI governance proposals out there, including over 600 bills in U.S. states alone, we need to focus on real harm, not just throw a blanket over research. And since AI is a global thing, our regulations should align with international standards as much as possible.

Looking Down the Road

AI is already making waves, from the tools you use every day—like Google Search, Translate, Maps, Gmail, and YouTube—to tackling big societal challenges. It's not just a tech breakthrough; it's a breakthrough in making breakthroughs happen faster.

Take Google DeepMind's AlphaFold, for example. It's predicted the 3D shapes of nearly all known proteins and how they interact. Or how about using AI to predict floods up to seven days in advance, saving lives for 460 million people in 80 countries? And then there's mapping the pathways of neurons in the human brain, uncovering new structures and helping us understand how we think, learn, and remember.

AI can keep driving these kinds of breakthroughs if we keep our eyes on the prize—its long-term potential.

That means staying consistent, thoughtful, and working together, always keeping in mind the benefits we can all enjoy if we get this right.