HR Compliance in an AI World with John Rood

Lucinda explores the critical intersection of AI governance and compliance within modern organisations with special guest John Rude, founder of Perceptual, who emphasises that as major regulations like the EU AI Act emerge, AI oversight must transition from a niche IT concern to a cross-functional responsibility involving HR, legal, and executive leadership.

They discuss how high-risk applications, such as recruitment and performance management, require robust documentation and ethical frameworks to mitigate bias and liability, providing a wake up call to action for businesses to implement internal AI policies and tiered literacy training to navigate the rapid evolution of technology safely and strategically.

KEY TAKEAWAYS

Organisations must prepare for the EU AI Act, which is set to establish a global standard similar to GDPR. It categorises AI uses by risk, with high-risk areas requiring extensive documentation and management systems.

While AI governance often lands on the desks of HR or IT, it must be an organisation-wide effort. Restricting governance to a single department can lead to "Shadow AI," where employees use tools without oversight, increasing liability and bias risks.

Effective governance requires a tiered approach to training. Executives need to understand strategic risk, middle managers need function-specific context, and all employees require a baseline of AI literacy to avoid basic security pitfalls.

The absolute minimum requirement for any organisation today is an Internal AI Use Policy. This document acts as the first line of defence, defining how employees can and cannot interact with AI tools to protect company assets.

BEST MOMENTS

"If we say governance just belongs only in HR, or only in information security, or only in IT, it doesn't end up working... the policies we create to put governance into place have to filter throughout the entire organisation."

"The EU often times sets a global standard based both on their desire to act quickly on new items... and the expansiveness with which they're willing to regulate."

"It's the potential risk to individuals to over-benefit some and disadvantage the disadvantaged... it’s that kind of impact on humans if not used with great ethics."

"Every organisation needs an internal AI use policy, and if you don't have it, that really is in my mind like an emergency."