Guardrails and Guidelines: Top Talent Leaders Weigh in on How to Use AI Safely

Whenever a major new technology is introduced, the excitement about its potential is tempered by concerns over the problems it might create. And that’s never been truer than with AI. While AI could revolutionize talent acquisition — relieving recruiters of mundane tasks to focus on work they love — it also poses risks. 

The good news is that you don’t have to figure out how to navigate this all on your own. At Talent Connect, we brought together four talent leaders for a breakout session called Recruiting with AI: Guardrails, Guidelines, and Ethical Considerations to share ideas about how to use GAI safely. 

Chris Hoyt, the president of CareerXroads and a talent acquisition expert, led a conversation with Melissa Thompson, the global head of talent acquisition for the Ford Motor Company; Mark Fuell, the global head of talent attraction for EY; and Marie Artim, the vice president of global talent acquisition for Enterprise Mobility

Together, they discussed how to tap into AI’s power in an ethical and responsible way. Let’s look at what they had to say. 

Everyone’s in a different place in their AI journey

In many ways, the leaders LinkedIn invited were a perfect microcosm of where companies are in their AI journeys. Marie — whose organization encompasses Enterprise, National, and Alamo rental car brands, as well as technology solutions and more — said that Enterprise Mobility “was a cautious organization, moving slowly into this space.” She looked out upon the audience and added, “I hope I’m not alone.”

Meanwhile, Mark said that EY is relatively advanced in its implementation of AI. Over the last two years, the professional services firm has invested more than $1 billion in its own AI platform, which is called ey.ai. “So, pretty much everyone,” he said, “is expected to have some foundational knowledge in AI and how it works.” 

Ford is in a similar spot, having created its own internal large language model called Mach One. “Yeah,” Melissa joked, “got to have those car analogies.” She said that her TA team had also recently turned on new candidate relationship management (CRM) software with AI technology built into it. “And what I find really interesting is, some of the recruiters have immediately picked it up,” she said. “They got that this is something that is going to truly help.” 

Use AI responsibly and effectively, even if it means taking the slow lane

Once each panelist identified where they are with AI, Chris asked how they’re implementing the technology responsibly. For Marie, the answer was simple. “We’ve always taken a very conservative approach to anything tied to selection in any way,” she explained. The company looks at what’s at stake for everyone, including candidates, employees, and hiring managers. But despite the measured approach, Marie said her talent teams were eager to embrace AI. “They can’t wait to use it,” she said. 

Because EY is further along with AI, they have different concerns. They’re using AI for automating processes, prioritizing workloads, and improving productivity, Mark explained. “But where we draw the line is that we will not allow AI to make any decisions,” he said. That applies not just to hiring decisions, but to screening decisions at the top of the funnel. “The candidates we receive,” he said, “are actually reviewed by a human being.” 

EY has its own ethical AI practice, which has helped its risk management team create an exhaustive list of questions the company needs to ask every vendor before signing a contract with them. 

Like EY, Ford doesn’t allow AI to make any decisions. And Melissa said that even though her new AI-powered CRM helps “streamline the mass we have at the top of the funnel,” getting the system up and running took time, due to Ford’s fairly cautious approach. “It took time,” Melissa said, “to get legal, data privacy, and others to understand how the tool works and that AI was not making decisions.” 

How to innovate with AI (hint: consider offering it the occasional bribe) 

Of course, the reason companies want to adopt AI is because of its huge upsides — which is why Chris asked each panelist to share the most innovative way they’re using AI. 

“Probably the thing that we have used it most frequently for, which I absolutely love,” Melissa said, “is OKRs [objectives and key results].” She explained that people often have a hard time writing “a key result that is actually measurable.” But with AI, you can enter an objective and ask it to “make this thing measurable.” It may take a few tries, but Melissa suggested that you keep refining the prompt until “you actually end up with something that’s measurable, realistic, and can be completed.” 

She’s also used AI to help her create a metric for “quality of slate,” so she can measure the effectiveness of her recruiters. Melissa went back and forth with AI until she arrived at three measures — time to source, conversion rate from recruiter screening to interview, and a satisfaction result from a hiring manager survey (“Did you get a quality slate?”) — that will be blended into a single quality-of-slate metric. 

Mark’s team at EY, meanwhile, has created a playbook of prompts for all team members to use. “It’s peer-reviewed and everybody has access to the prompt playbook,” he said, “and they’re able to make recommendations and suggestions as to how to improve it.” 

Mark said that he’s seen a few wild things that he didn’t think would work, but did. “We had one person who was looking for talking points in a generic way, and the AI came out with some talking points,” Mark said, “and then this person followed up with, ‘If I gave you $1,000, could you give me more interesting talking points?’ And it did!”

The ethical considerations talent teams should weigh — and candidates should too 

Chris asked the panel, “What ethical considerations are you weighing as you develop a strategy for AI implementation?” 

Mark noted that there are two sides to EY’s ethical considerations — that of the company and that of the candidates. For the company, he said, it goes back to decision-making, but also that the company wants to be fair and transparent about how it uses AI in the hiring process. 

As for candidates: “We do a ton of early-career hiring, and we’re seeing a lot of creativity on that side,” Mark said, “so how do we get in front of that? What are the things that we can strategically be ready to talk about, set in place, and discuss?” 

“We had a quarterly TA review with our IT leadership team,” Melissa jumped in, “and we had a really robust discussion around this.” One of the things that came out of their discussion was that Ford wanted future employees to use ChatGPT. “So, how do we figure out a way to allow candidates to use it in some instances?” she asked. “That way, we can see whether they’re good at creating prompts.” 
But in terms of interviews, she added, there are times when it’s appropriate to tell a candidate: “For this one, we want you to just use your brainpower and answer the question.” 

Final thoughts: Change management is key (and a good pilot doesn’t hurt) 

Toward the end of the discussion, Chris asked the three talent leaders how they handled change management when implementing AI

“One of the first things I like to say to my team is, ‘You’re not going to adopt this because I tell you to,’” Melissa said. “‘You’re going to adopt it because you can see how it adds value.” 

Mark chimed in, saying that to create effective change management, support from leadership is key. Leaders need to be very clear about why they’re implementing AI and what their objectives are. “I’m also a big fan of pilots and phased rollouts,” he said. “You learn from your mistakes and move on to the next thing, which is a very important element of the change management process.” 

“At Enterprise Mobility, we definitely love a pilot,” Marie said. When the company rolls a new technology, she said, it usually starts with a pilot. And then the people who participate in the pilot become champions for change. 

“Employees have to see the value in the technology, but they won’t see the value if they don’t try it,” she explained. “They need someone to show them, ‘This is how it worked for me.’ If we tell them, they don’t always buy in. But if their peer tells them, they do.”

Uncategorised