Embracing the future of HR by becoming an AI-first enterprise

If you’re a CHRO, you probably can’t read a news article or blog these days without hearing about agentic AI. It’s everywhere. The benefits of AI agents are impressive, but for some CHROs, AI agents are still a little “sci-fi” and traditional AI feels more appropriate for their business.

Our IBM HR team has experienced both points of view. We have undergone a complete digital and cultural transformation thanks to AI—and we started in 2017. Because this changes started long before many companies were even thinking about the technology, we learned a lot along the way. And yes, we made mistakes.

If we could start over, we’d do some things differently. Keep reading and see the things we’ve outlined to help the CHROs who are just starting out or those who are trying to decide the next steps in their AI journey.

It’s no secret that becoming an AI-first enterprise has become a necessity. By using AI, we can unlock new levels of productivity, efficiency and employee experiences—all things employees are expecting in the workplace.

This kind of digital transformation is not without its challenges. It’s like renovating your home while still living in it. This kind of transformation requires careful planning, strategic execution and a willingness to change. That willingness isn’t only about accepting new technology; teams must simultaneously address company culture, service delivery and internal processes or the technology will be a nonstarter.

We’re often asked what it means to be an AI-first enterprise. At IBM, AI-first means to harness the power of artificial intelligence to enhance our capabilities, encourage creativity and empower our teams. 

It’s about creating a culture where AI is seamlessly integrated into our daily operations and workflows, augmenting human capabilities rather than replacing them. AI takes away the drudgery of work, allowing us to do those uniquely human, higher-value tasks that only we can do.

When we started in 2017, did we see a complete vision of where we would be in 2025? No. It wasn’t a big, strategic decision. Not at all. Instead, the impetus was necessity, a confluence of challenges coming together to create demand for a new solution—this situation should ring true for many companies. 

HR’s compliance work was growing increasingly complex because of new laws and regulations around the world and the team was struggling to keep up with the rate and pace of change. At the same time, consumers were coming to rely on apps offering customization, personalization and convenience and employees were bringing those expectations to work. 

Normally we would tackle issues like these by throwing more money at the problem; that was no longer an option due to a consistently shrinking budget. We knew that we had to work differently.

AI became the obvious solution. When we rolled out AskHR—an HR question and answer chatbot—in 2017, it set us on the path of being Client Zero. During this journey, we tested and implemented IBM’s AI and automation technology first, showing the power and potential for both IBM and clients alike.

The problem was that none of our employees used AskHR in 2017. Why would they when they could still walk down the hall to talk to an HR Partner, call the 1-800 number or send an email and get an answer in 24 hours?

We kept working to improve the technology and in 2018, we made a bold move. We shut down the HR phone number and email address for all employees and discontinued HR Partner support for 21,000 first-line managers literally overnight. AskHR became the digital-first entrance into HR.

Why did we do it? We had to force that culture change to get our employees to start thinking and acting differently. At first, it was really challenging. Our NPS plummeted to -35. But then IBMers realized that with a chatbot that was always “on”, they got the answers they needed in seconds.

As the chatbot learned and got smarter, our NPS began to rise. We added more functionality, along with the ability to do transactions. AskHR morphed into a digital assistant that allowed managers to transfer employees to another manager or help them initiate the quarterly promotions process. And it was all done right in AskHR with just a few clicks.

In 2024, AskHR handled more than 11.5 million interactions; 94% of those interactions were contained within the platform. That means out of all the questions that were asked, only 6% needed to be routed outside of AskHR to a specialized HR Partner for assistance.

The current NPS is +74, so we have come a long way from that -35 score. There are almost 90 automations embedded into AskHR and more are on the way. Because of these automations, managers can do HR transactions 75% faster than before.

AskHR recently moved to its new home on IBM watsonx Orchestrate, our generative AI platform. We brought our AI agents along, so now IBMers can find what they need in one place. That is the beauty of agentic AI—the orchestration layer.

Now IBMers can just go to AskHR instead of multiple platforms. They can:

• Have a chat about careers
• Onboard a new employee
• View, promote, schedule and manage learning classes
• Get help with the quarterly promotions process

All in one place. This agentic capability allows us to unlock new levels of productivity, efficiency and employee experience for IBMers worldwide.

As mentioned in the beginning, we have learned a lot along our AI journey and there are a few things we would have done differently. Here are the lessons that we learned and the best guidance we’d give if starting out fresh.

Instead of going “big bang” like we did for AskHR, find a small pain point that wears on your workforce. Automate that, make it great and then scale. 

You will learn a lot on your AI journey, too. If a pilot is not working, don’t be afraid to abandon it. Try something else. Be agile.

Yes, HR has great ideas, but so do employees and managers. Give them a platform to submit their ideas and then implement the best ones. When employees see their ideas in action, you will build advocates quickly.

Just because the company down the street is doing something and it works for them, it doesn’t mean it is going to work for you. Whatever AI you implement, it must be in service of your business and your employees, not what’s “cool” or what everyone else is doing.

You’ve heard about our journey and guidance on how to get started. Here is some additional high-level advice on what to look out for as you implement AI in HR.

Today we are talking about AI, tomorrow perhaps we are planning for quantum. The half-life of skills continues to shrink; our own research shows a range of 2.5–5 years for most skills in the tech and HR fields.

We must ensure that our organizations are equipped to continue navigating the AI transformation—or any future transformations—successfully. That means to have the right people with the right skills.

Do you plan to reskill and upskill your workforce for AI? Are you thinking about how to redesign jobs for the higher-value work your teams will be doing when AI replaces those routine tasks, particularly for your entry-level positions? And what are your teams doing with the time that they’re saving? Without clear direction, that time is wasted.

HR is in a position to navigate the “natural tensions” of welcoming AI responsibly and successfully into a workplace. Part of an organization might want to blindly implement AI without proper attention to ethics or workflows. Another part might want to take a more conservative approach that could leave the organization behind competitors. HR sits at this unique intersection of understanding those points of view and finding a path forward.

While IBM HR abides by IBM’s pillars of AI Ethics for all our new AI implementations, there are concerns about the continued ethical implications of AI in the market. It is imperative that we ensure others are using it in a way that is fair, transparent and respectful to customers and employees alike.

Simply put: no pursuit of innovation can ignore the needs of the workforce. Any transformation, AI-related or not, will be successfully adopted only if it is empowering employees for success.

One of the most significant opportunities for AI in HR is in employee experience. With AI, HR professionals can flourish as employee experience champions. Through a recent study from the IBM Institute for Business Value, we learned that companies that prioritize employee experiences grow their revenue an average of 31% faster than others.

As we looked to define the AI experience we wanted for IBMers, the mantra we developed for our decision making was, “Eliminate, simplify, automate”. Here’s how it works:

To build an effective business case for AI, the people proposing its use must first illustrate why a certain policy or process needs to exist at all. Then, if it passes that test, the process needs to be simplified as much as possible. This process ensures that we have only the best processes ready for AI and automation.

Here’s a great example from our own team: when reviewing our company leave policies, we realized we had one for employees taking part in the Olympic Games. It wasn’t necessary to have a detailed, one-off process for the few IBMers that it might affect every couple of years, so we eliminated that outright. In fact, we consolidated over 25 distinct types of leaves down to one.

In the last four years, we have seen a 40% reduction in the HR operating budget. In 2024 alone, AskHR successfully handled more than 11.5 million interactions and completed over one million transactions.

This increased our productivity and saved both managers and employees precious time in their day, time that they now use on higher-value work. Our HR team is proud that our Client Zero work contributed to the USD 3.5 billion in productivity savings (against a USD 2 billion target) that IBM realized in 2024.

As we look to the future, we see immense potential for AI agents in HR. When we recently made the shift to agentic AI with our newest release of AskHR on IBM watsonx Orchestrate, for example, it was an exciting day for our team. There are so many other things that we can do with the employee experience now that we have agentic AI capabilities.

With AI 1.0, AskHR was a simple AI chatbot: you asked a question with a prompt, you got a pre-curated answer in return. Agentic AI is the next frontier. A combination of the right technology, agent autonomy and human oversight is the key to making AI agents work for HR, our employees and our clients. 

These agents will free up HR professionals even more to focus on strategic, high-value work. AI-driven analytics and insights will help us make better decisions, predict talent gaps, identify opportunities for growth and better support our employees. 

Whether you’re ready to take this next step in adopting AI agents or are still dipping your toes in with traditional AI, one thing is clear: you can’t wait to get started.