As DeepSeek makes headlines, how should HR navigate the latest AI tools?

'With these technologies, the part that’s a little scary is the fact that they are super technical,' says academic, offering tips for HR

As DeepSeek makes headlines, how should HR navigate the latest AI tools?

Just as many business leaders are getting comfortable with generative AI (genAI) models like ChatGPT and trusting employees to use the tools, a new product has thrown yet another wrench into the machine and raised more questions about the future of AI in the workplace.

This week’s release of DeepSeek, an open-source genAI model that apparently rivals ChatGPT at a fraction of the price, has upset the current status quo and shown that American companies like Meta and Anthropic do not hold the monopoly on innovation.

The unveiling of DeepSeek has underlined a major challenge for employers: how do you implement and monitor workplace tech strategies when the tools themselves are constantly changing?

Kevin Lee, assistant professor of organizational behavior and human resources at the Sauder School of Business, sheds light on the implications of these disruptions and provides guidance for navigating this transformative era.

Choosing right tools in crowded market

With the emergence of DeepSeek – and other said to be close behind – the challenges for HR professionals are mounting in assessing the capabilities and how they apply to the workforce.

Lee highlights the importance of tailoring AI adoption to organizational needs – and HR not doing it on its own.

"With these technologies, the part that’s a little scary is the fact that they are super technical,” he says. “They're a little bit beyond our everyday knowledge about how they function, or what kind of effect it might have.”

Experts in the field have reported that DeepSeek’s model rivals OpenAI’s ChatGPT and even exceeds its performance in some areas. However, the open-source nature of DeepSeek means it is less secure than the more expensive proprietary models designed by Silicon Valley giants, adding even more variables and concerns for organizations using the tools.

This is the moment for collaboration, Lee says, when different departments and divisions within organizations should be working together and sharing knowledge. He advises HR professionals to partner with IT teams and establish committees that bring together technical and organizational expertise.

“Popping the hood of these technologies, sitting down with people who are more knowledgeable about them, getting a better understanding of what the landscape looks like – I think those things are really important for the kind of deep engagement required,” Lee says.

It’s about asking: “Based on the tools that we have out there, what is actually well suited for the organization? What are the pain points within our organization? How can we sort of change our organization, based on technologies?” he says.

Training employees amid constant change

The fast-paced development of AI tools poses another significant challenge: training employees effectively while contending with tools that quickly become outdated or replaced by newer, more advanced models.

Consulting with industry experts and attending conferences is important for being up to date on industry developments, but Lee also stresses the importance of regular engagement between HR and employees who are using the tools.

“Just sit down with… employees regularly. Maybe have some sort of regular sessions or meetings, maybe appoint a committee of employees,” he says.

“Check in with them regularly as representatives of the broader employee body. What exactly are the pain points within the workflow, based on the knowledge that you've been gaining as an HR professional, about these different kinds of technological solutions. At any given moment, what actually might be useful?”

Maximizing current talent and hiring for AI skills

As generative AI becomes increasingly integral to workplace operations, organizations must adjust their hiring strategies, and Lee emphasizes the need for domain expertise in AI and related technologies.

When looking at current employees, he recommends identifying individuals who are already showing themselves to be interested in the new technology, who have been experimenting with it, working with it, learning about it.

These individuals can be valuable point-people for the rest of the workforce, he says, as key people to help lead the way, “with regard to what do we do with AI, what is going on out there, and what exactly would actually be specifically good for our organization.”

When hiring for roles with more technical requirements, Lee cautions against hiring applicants with self- or online-taught skills and certifications, “because there is, of course, the difference between people who know the technology inside out and can understand it from the ground up, as opposed to other people who don't really have the technical base,” he explains.

“Maybe you have an ad hoc knowledge of AI specifically, but maybe missing some important things about some of the technical aspects of what these technologies entail, and by extension their social consequences,”

Going forward, creating committees made up of people with domain expertise is advisable, says Lee, who know well the sort of work being done within the organization. And for roles that don’t require dedicated AI specialists, HR professionals can prioritize candidates with basic familiarity or a willingness to learn AI tools, he says.

The DeepSeek disruption: a game-changing competitor

While the rapid evolution of this technology can feel overwhelming, it’s not the first time we’ve experienced disruption, and now is not the time to worry, says Lee.

“Every time we have new technology come in, it's never quite as bad as people think it's going to be. It's never quite as apocalyptic.”

On the contrary, he says this moment of transition is the time to be collaborating and learning.

"This is a time of experimentation. None of us particularly know what the future is going to look like. None of us particularly know how exactly these technologies will unfold. It’s an interesting time to be playing with these technologies, experimenting with them, kind of innovating and imagining."

Latest stories