AI governance “moving from ICT teams to Boardrooms” asserts BSI

Last year, the BSI highlighted a concerning governance gap when it came to AI. Now, research carried out by the global professional services company shows that, amid rapid growth in adverts for chief AI officers or firms announcing high-profile appointments, AI governance has shifted from sitting within ICT or strategy functions to being a Board-level and executive concern.

The research paper explores growing recognition that AI is not just another tool, but instead represents a profound business realignment, with more than 20,000 CXO job titles on LinkedIn that include AI or Artificial Intelligence. Despite this, accountability for AI remains fragmented in most organisations, with responsibility for AI strategy, risk management and oversight often spread across roles and functions, in turn creating ambiguity about ownership, liability and decision-making (and particularly so when things go wrong).

Identifying a range of accountability models, the BSI’s report suggests that no single ‘best’ approach has yet emerged. Currently, some firms have chief AI officers, others exhibit expanded CIO/CDO remits, federated governance councils or business unit ownership.

Faster movement 

Large enterprises and high-risk sectors such as financial services are moving faster towards formalised accountability and governance, while SMEs tend to adopt more fluid approaches that are capability-led in style.

The research paper concludes that good governance is less about someone having a designated title, such as chief AI officer, and more about that person or team being empowered, having cross-functional reach and the resources and ability to effectively operationalise responsible AI.

Further, the BSI notes that AI maturity drives governance maturity and calls for organisations to prioritise establishing more formal oversight as they move from experimentation to widespread operationalising of AI usage.

Business-level accountability

Tim McGarr, global head of AI market development and partnerships at the BSI, commented: “AI accountability is more than a title. It’s the expectation that organisations demonstrate credible and operationalised business-level accountability. As AI scales across enterprises or into commercial models, formal accountability, controls and named owners become critical.”

McGarr added: “Good AI governance is fast becoming a leadership and trust issue, with expectations from regulators, employees and the public alike that organisations can demonstrate credible and auditable responsibility for AI outcomes. The time is now to build your AI accountability approach. One that can grow with your business to embrace AI responsibly.”

Prior research results at the BSI found that fewer than a quarter (24%, in fact) of organisations surveyed reported having an AI governance programme in place, while under half said its usage was controlled by formal processes.

*Further information is available online at www.bsigroup.com