AI sprawl is a talent problem: Ownership must belong to HR leaders says EY

AI sprawl and unsanctioned AI use is no longer a fringe concern for IT. It has become a practical, people-based challenge for HR leaders impacting skills, trust, productivity, governance, and talent advantage.

This scale of ‘shadow AI’ use ranged from 23% in Government and public sectors, to 58% in wealth and asset management, according to EY’s recent Work Reimagined survey.

Employees using unsanctioned AI tools, often at their own expense, for work is a symptom of ‘AI sprawl’ – the uncontrolled proliferation of AI tools, agents, models, and data projects across an organization without centralized governance or oversight.

As organizations press ahead with official AI implementation and employees stick with other, preferred tools, HR leaders are faced with a significant challenge that requires them to act now.

Speaking to UNLEASH, EY Global People Consulting Leader, Kim Billeter, explains that to realize productivity or other returns from AI, organizations must reframe and focus on three key drivers: mindset, skillet and toolset.

Get policy and tools right first before attempting to take on AI sprawl

Both AI sprawl and shadow AI use can be addressed by ensuring communication with employees is transparent.

According to EY, employees must be conditioned for the journey and destination the organization is on (mindset), the upskilling of the workforce (skillset) and the governance over which tools are part of the strategy (toolset).

“We've got to really change the operating model, change how people are looking at the skillset. Governance and policy are two things, from an AI sprawl perspective, that we've got to really hit hard,” Billeter states.

She tells UNLEASH that people leaders should focus on adopting the right tools and implementing relevant policy first before taking on sprawl, as “that's going to be really hard to govern.”

You’ve got to have policy, and if these programs run well, people will start to see how they can get efficiency, driving their career trajectory by using the tools an organization wants.You’ve got to have policy, and if these programs run well, people will start to see how they can get efficiency, driving their career trajectory by using the tools an organization wants.

“There are way too many organizations that are just throwing technology at people and hoping for the best that they're going to utilize it well.”

It’s easier said than done; Billeter believes many organizations are somewhere between having policy “locked down” on what tools employees have access to and “the Wild West” where employees are able to use unsanctioned tools freely.

Critical factors to achieving effective governance of AI use include getting the tools right, giving people enough direction, and having an AI-focused employee value proposition in place.

Organizations and people leaders must not be in thrall to agility either, Billeter says, because “as fast as this is moving, it can't change your mind every two minutes on what the next startup or the next big thing is.”

Harness the power of managers and internal influencers to tackle shadow AI

Aliqua magna fugiat esse eiusmod sint aute irure voluptate cupidatat aute adipisicing velit. Esse incididunt magna et velit do id commodo quis consequat voluptate. Velit consectetur cupidatat sint ad eu amet et ex aute ipsum duis exercitation labore anim. Sint elit tempor sunt consequat aliquip consequat consectetur.

Harness the power of mangers and internal influencers to tackle shadow AI

HR leaders addressing unsanctioned AI use within the organization must take the temperature of employee sentiment: how do they feel about the tools they are allowed to use? What tools do they use outside of work? Which tools would be more effective to approve for work?

Billeter recommends “finding the influencers in your organization and have them push and pull information on how to use tools” and on what tools employees are using and how.

Traditional employee engagement methods like pulse surveys can also be effective here, Billeter says, as they are anonymous and offer a space for employees to speak freely.

“Couple that with good communication on where you are going on strategy and where you saw success, expectations on and incentivization for employees… there's got to be a broad strategy to get there, not one avenue that's just not going to solve it.”

Managers can also play a significant role in changing employee behaviors, but it requires adjustment on their part to help drive engagement of policy and governance: “It is a constant change management battle, and behavior change at scale, especially for managers and leaders.”

“You have to start with both a leadership down in terms of strategy and a grounded up campaign to drive activity in the right way, connected to HR programs and the upskilling efforts.”

Using AI as a talent advantage can be the ‘second moment’ for HR leaders

EY’s Work Reimagined Survey found that employees with 81 or more hours of AI training often see major improvements to their skills and report an average productivity gain of 14 hours per week.

However, cultivating highly skilled employees also increases the risk of them leaving the organization; this group of workers with 81 of more hours of training are 55% more likely to seek new opportunities.

That retention risk can be mitigated by focusing on how employees are able to realize the efficiencies AI can bring; allowing for more time to spend on wellness and family, build new skills or work on other areas of professional development.

Making that return a tangible benefit or development for employees will help address attrition as the battle for AI skills grows.

"If your organization isn't advancing now that they're hooked, they will go to more innovative organizations – they now have the skillsets to do it,” Billeter warns.

The talent advantage plus AI advantage is where the leading companies are going to win, but you've got to be able to paint that picture for folks.The talent advantage plus AI advantage is where the leading companies are going to win, but you've got to be able to paint that picture for folks.

“Part of that is use cases and you need to do use cases in HR as a proof point. HR is a wonderful place to start with AI in terms of automating the processes and deploying agents.”

Ultimately, Billeter believes this convergence of talent and AI is “the second moment” for HR leaders, following the “arrival of the CHRO” during the COVID-19 pandemic.

The job in front of HR leaders now is to address AI sprawl and shadow AI use by owning the strategic workforce plan.