Corporate Risk Radar | Series 2, Episode 2 | Navigating regulation and risk in AI

In this second episode of the Corporate Risk Radar Podcast Series Two, host Rebecca Kelly is joined by Chris Williams, Partner in intellectual property (IP) and technology disputes, and Heidi Watson, Partner in the employment team. They delve into developments around AI regulation and implications for businesses, as over a third of senior leaders say they feel unprepared for AI driven disruption, and 58% say regulation is already a significant burden.

Williams kicks off by outlining the UK’s current “flexible, pro-innovation approach” to AI, consisting of a set of principles covering topics such as IP, data protection, consumer law and competition law. While not legally binding in their present form, with a new government now in place, these look set to be codified in the near future. 

From an HR perspective, Watson discusses AI use cases in recruitment, attendance, and dismissal decisions, where the potential for discrimination and bias are key risks. Legal challenges also arise when HR professionals can’t explain how AI decisions are made, and Clyde & Co’s primary focus has been raising awareness of these issues so clients can take steps to mitigate them. Transparency is key, as Williams notes: “If you're using something, you need to stand behind it in terms of potential accountability.”

Naturally, legal teams are keeping an eye on emerging cases and, while none have yet gone to full hearing, Watson has seen many being raised and settled, including potentially racially discriminatory facial recognition technology at Uber Eats. The HR community is “waiting to see how these issues are being dealt with around the world,” she says.

Among IP cases, a common theme is lack of understanding on how data can be used for training AI models and, as Williams explains, “a steep learning curve for many in terms of what they can and cannot do with the information in their possession”. On the other side of the equation, firms are grappling with the challenge of determining authorship of AI outputs and protecting these when copyright laws may not apply.    

To stay ahead of regulations, Williams recommends a “layered approach,” addressing factors both within and beyond a company’s control. Internally, businesses can implement an AI policy, designate an AI oversight role, and prepare for stricter regulatory frameworks. Externally, they must carefully assess the risks of sharing sensitive and confidential data with third parties, where control over its use may be lost.

Ending on a positive note, Watson says there are “incredible opportunities here for HR” when balanced with managing the risks by understanding the software and potential for bias, and the critical role of people. Williams echoes the sentiment, stressing the importance of preparation via an AI policy, plus effective communication and training. “If people understand the risks and why these policies are there, they're much more likely to engage” he concludes.
  
Look out for the next episode of our Corporate Risk Radar Podcast, which will focus on geopolitical risks. Subscribe to the Clyde & Co podcast channel for more information.