The current “vibe” in the United States, United Kingdom and even the European Union appears to be one of relative hesitancy toward new regulation and this may leave governments on the backfoot as agentic artificial intelligence advances, ZoomInfo’s chief strategist for privacy and artificial intelligence told MLex.
The current “vibe” in the United States, United Kingdom and even the European Union appears to be one of relative hesitancy toward new regulation and this may leave governments on the backfoot as agentic artificial intelligence advances, ZoomInfo’s chief strategist for privacy and artificial intelligence told MLex. “I'm a bit of a pessimist here, where I think … we're going to have very little new AI regulation for a year or two,” the US software company’s Simon McDougall said on the sidelines of a privacy conference* today. “But with agentic AI in particular, some bad things are going to happen. And that will drive reactive regulation.”
Proponents of agentic AI say the technology can serve as a collaborator of sorts, completing tasks and solving problems autonomously (see here).
“When you have a technology which can self-improve, [that] can analyze itself and write its own code, and now we're allowing it to engage with the outside world, it can sometimes be very persuasive,” said McDougall, who is also on the IAPP board of directors and has been deputy commissioner in the UK's Information Commissioner’s Office. “It can sometimes be quite deceptive … This is not to be taken lightly.”
Former US President Joe Biden and Former UK Prime Minister Rishi Sunak focused on the risks that advanced AI systems pose, and prioritized studying and assessing these models through specialized AI safety institutes.
These institutes are still in place, though the UK’s recently rebranded to focus on AI security (see here). Meanwhile, US President Donald Trump and UK Prime Minister Keir Starmer are emphasizing AI’s potential as an economic driver and the importance of innovation over regulation.
The EU, by contrast, passed the world’s first cross-industry AI-governing law, the EU AI Act, in March of last year (see here). It’s still in the process of rolling out.
“I think that if the EU AI Act was being drafted right now, it would look very different to the one that was passed,” McDougall said. “I think there is definitely a notion of trying to think through how they can make the AI environment in the EU better for innovation.”
McDougall said, however, that the current “vibe” is “to me, a bit of an overcorrection.”
The US and UK still need specialized, credible institutes to test AI models, he said, and “what they are called doesn’t really matter too much.”
Companies are still concerned with their reputations, McDougall said, which can inspire care and caution even in the absence of regulation.
“Even though regulation per say is no longer fashionable … brand and reputation risk is very real,” he said.
*IAPP Global Privacy Summit 2025, Washington, DC, April 22-24, 2025.
Please e-mail editors@mlex.com to contact the editorial staff regarding this story, or to submit the names of lawyers and advisers.