Microsoft and OpenAI are dominating headlines, but which AI platforms are actually transforming HR?
Microsoft and OpenAI are reshaping one of the most important alliances in artificial intelligence, loosening exclusivity terms while insisting their relationship remains firmly intact.
In a statement released today, the companies said they are entering the “next phase” of their partnership, with Microsoft remaining OpenAI’s primary cloud partner and OpenAI products continuing to launch first on Azure. At the same time, OpenAI will now be able to sell products across other cloud providers, while Microsoft’s access to OpenAI intellectual property will continue through 2032 on a non-exclusive basis.
The move has fueled fresh debate over whether the AI market is entering a more competitive phase, with enterprise buyers potentially gaining more choice beyond a handful of dominant names.
But inside HR departments, the reaction is likely to be more pragmatic than dramatic.
That’s according to Ben Eubanks, Chief Research Officer at Lighthouse Research & Advisory, who says many HR executives are less interested in Silicon Valley power shifts than in whether their existing tools can save time, improve decision-making, and stay compliant.
“When I think about the times that I’m talking to HR leaders, I would bet that the majority of them don’t actually know that OpenAI models were the ones that power Copilot,” Eubanks said. “The average HR person doesn’t have a lot of free time to go and use three or four different tools. They’re going to use the one that’s closest at hand.”
That means the biggest winners in HR may not be the most advanced platforms. They may be the ones already embedded in daily workflows. Companies running on Microsoft 365 are likely to default to Copilot. Organizations built around Google Workspace may lean toward Gemini. Familiarity often beats novelty.
What does HR really want in their AI platforms?
While AI vendors compete over model speed and technical benchmarks, Eubanks says HR teams usually evaluate tools through a much narrower lens: can this solve a problem I already have?
“They’re not even talking as much about the platform itself as they are the use cases and examples,” he said. “They are hearing all these headlines telling them that AI is the next wave and it solves all problems. That’s too vague for them.”
That practical focus helps explain why different AI platforms are finding different footholds in HR.
Read more: AI was supposed to elevate HR. What if it does the opposite?
According to Eubanks, generalist HR teams often use Microsoft Copilot because it’s already integrated with Outlook, Word, Excel, and Teams. Recruiters and individual users frequently turn to OpenAI ChatGPT, sometimes paying personally for subscriptions before their employers formally adopt a company-wide tool. Meanwhile, users with more advanced needs may experiment with more niche platforms.
“I will tell you that the people I know that are in analytics or compensation, those people swear by Claude until they’re blue in the face,” Eubanks said.
That reflects a growing split in the market. Mainstream users want convenience and integration. Power users want stronger writing, reasoning or data analysis. Eubanks also pointed to specialized tools such as Julius AI, which his team tested for analyzing large datasets. That kind of specialized software could appeal to people analytics teams that need more than a standard chatbot.
The real obstacle isn’t the software
Even as the platform race intensifies, Eubanks says the biggest barriers to adoption remain internal. He cited recent Lighthouse research surveying nearly 1,000 talent acquisition leaders which found that 80% said they had faced issues adopting AI in talent processes. Of those challenges, about 80% were tied not to the technology itself, but to people, processes, work design, and skills gaps.
“The challenge with this conversation on AI is it’s treated like you turn on a light switch and it just goes and does everything,” he said. “If it’s going to add the most value, we’re going to have to think about how work is going to change.”
Security concerns are another brake on progress. Many organizations still restrict what employee data can be entered into outside AI tools, limiting some of HR’s most promising use cases. But those guardrails often reflect legitimate fears about privacy breaches, leaked information and misuse.
Eubanks warned that some organizations are moving too quickly, plugging AI into recruiting workflows without fully thinking through the risks of providing candidates’ personal data to a third-party platform.
“You would never print a folder of resumes and hand them to some random person on the street and say, ‘Bring those back to me once you’ve ranked them,’” he said. “And they’re doing the digital version of that.”
That, he argues, is where many companies go wrong. The technology may be powerful, but convenience can create reckless behavior if guardrails are missing.
“Someone breaks something, someone hacks something, and that stuff gets out in the wrong way,” Eubanks said.
Still, Eubanks believes hesitation carries its own risks. Teams that never experiment may fall behind those learning how to use AI thoughtfully, with clear guardrails, and a sharp focus on real business needs rather than novelty.
“There are still so many people that are just not quite sure,” Eubanks said. “‘I’m about to dip my toe in. Should I?’ Just go for it.”
That does not mean rushing headfirst into every new tool or handing sensitive decisions to algorithms. It means starting with contained, practical use cases, such as reducing repetitive admin work, improving training workflows, supporting recruiters with resume screening, or helping managers draft clearer communications, while keeping human oversight firmly in place.
His comments also point to a balancing act many organizations are now facing. Move too slowly, and teams may miss productivity gains and valuable learning time. Move too quickly, and companies can create privacy risks, weak processes, or poor employee experiences that are harder to reverse.
For companies watching the latest Microsoft and OpenAI headlines, Eubanks’ advice is to spend less time worrying about which platform wins and more time building the internal habits that make any tool successful: governance, training, smart experimentation, and clear accountability.
The longer-term opportunity, he argues, is not replacing human judgment but creating more space for it. If AI can handle some of the repetitive and administrative burden, HR teams can spend more time on culture, leadership, coaching, and workforce strategy.
“AI can’t replicate creativity, compassion, curiosity,” Eubanks said. “Those are innately human characteristics.”
