Threading the Needle in Defining AI’s Regulatory Boundaries

How does one tame the hyper-rapid advance of the disruptive technical beast?

While marketplace innovations have always played the proverbial “hare” to the regulatory “tortoises” of national governments when it comes to something like generative artificial intelligence (AI) and its unheralded commercial explosion — the part of innovation may be better played by something like a cheetah.

Still, that hasn’t stopped the 27-nation member bloc of the European Union (EU) from trying to strap the equivalent of rocket-powered roller skates onto their own tortoise to keep up with the pace of AI technology.

The EU’s regulatory AI act, the Artificial Intelligence Act (AIA), is already making progress and is expected to gradually go into effect starting next year.

So far, the EU’s swift-moving, frontrunner approach has left the U.S. in the dust. At least for now.

“It’s an interesting Rorschach to figure out, you know, what is important to the EU versus what is important to the United States,” Shaunt Sarkissian, founder and CEO at AI-ID, told PYMNTS CEO Karen Webster.

“If you look at all rules that come out of the EU, generally they tend to be very consumer privacy-oriented and less fixated on how this is going to be used in commerce,” he added.

Given that AI needs to be trained on data — lots and lots of data — the EU’s AIA may have the unintended consequence of shifting future innovations to other jurisdictions.

“If you make it difficult for models to be trained in the EU versus the U.S., well, where will the technology gravitate? Where it can grow the best — just like water, it will flow to wherever is most easily accessible,” Sarkissian said.

Get the researchPreparing for A Generative AI World

Finding a Balance Between Regulation and Innovation

That doesn’t mean the U.S. is poised to capture the lion’s share of AI’s growth flywheel.

Without clear regulations, companies may be hesitant to invest in AI research and development, which could slow down progress in the field.

That’s why collaboration between industry and regulators is seen as crucial for the growth of the AI industry.

Sarkissian explained that industry players should approach lawmakers with the attitude of, “We know this is new, we know it’s a little bit spooky, let’s work together on rules, laws, and regulations, and not just ask for forgiveness later, because that will help us grow as an industry.”

He added that, in his view, the first principles of AI regulation should be accountability and traceability, which are necessary to hold companies accountable for any negative consequences that may arise from their use of AI.

Key to Effective AI Regulation Is Defining What Is Being Regulated

AI’s ongoing commercialization will continue to grow and become ever more integrated into our daily lives, meaning the conversation around its regulation will only become more important.

“There are certain baseline guardrails that are going to be put in place, and that’s where the government can be very effective in my mind, as opposed to being prescriptive and saying you have to do these things,” Sarkissian said, something that he sees as being “90% regulating the use cases and maybe 10% regulating the technology.”

He gave as an example the dichotomy between a health inspector and a restaurant — where the inspector is responsible for ensuring that the restaurant meets certain criteria around cleanliness and process compliance, but whose role is another planet away from telling the chef what recipe to be using in the (clean) kitchen.

As PYMNTS has previously written, policymakers are currently contemplating several approaches to regulating AI, which broadly can be categorized across AI-specific regulations (EU AI Act), data-related regulations (GDPR, CCPA, COPPA), existing laws and legislation (antitrust and anti-discrimination law), and domain or sector-specific regulations (HIPAA and SR 11-7).

“There have to be rules that dictate what is an AI and what’s qualified as an AI or not … and then that filters into the technology around what needs to be tracked to make sure that it complies,” said Sarkissian. “There needs to be clear demarcation lines of what is considered generative and output-based AI and what is just running analytics at existing systems.

“In many ways, the EU legislation, even though it has some challenges, is forcing other nations to step up and expedite their own approaches,” he explained. “It’s an issue that people want solved and want done right.”