How AI Is Redefining Employee Recognition Programs

Artificial intelligence is revolutionizing employee recognition programs by making them more personalized and data-informed, spawning a variety of use cases beyond traditional applications. Employers are leveraging AI to analyze performance metrics, communication patterns, and employee preferences in order to deliver tailored recognition that resonates. By identifying unsung heroes and predicting when recognition can boost engagement or retention, AI empowers organizations to foster a culture of appreciation that is both strategic and scalable — transforming recognition from a reactive gesture into a proactive driver of workplace morale and performance.

Tom Libretto, president of Workhuman, spoke with SHRM at Workhuman Live  2025 about AI ushering in a new reality for recognition. 

SHRM Members-Only Toolkit: Managing Employee Recognition Programs

SHRM: What are some of the ways that AI is being used in recognition programs? 

Libretto: One very practical use case for AI is to evaluate the health of a recognition program — looking at the program and suggesting ways that it could be optimized for better outcomes. We have collected best practices related to how recognition and rewards programs are set up, launched, cultivated, and communicated. We have loaded that data into an LLM [large language model] and can serve those insights back to employers querying optimization opportunities in their own programs.   

The second way is more interesting. There’s an almost limitless runway in use cases where the recognition data itself is mined for rich, authentic, peer-based insider knowledge around who is doing what work in the organization, what skills are being invoked to do that work, and what they are being recognized for relative to the impact that work is having on the broader company.  

When enough volume of that data exists, it is ripe for mining with a recognition-optimized LLM and an agentic AI interface. You can ask all types of people-related queries in a conversational way. Who is showing leadership skills that is predictive of a promotion in the next 18 months? Who is the cultural flag bearer in this department? Who are the top 10 candidates that meet the skills criteria for an open position? Those talent and L&D [leadership and development] use cases are nearly limitless.          

SHRM: How can HR ensure that AI is being trained on quality recognition data? 

Libretto: Having the tooling within the recognition program technology can help. Our platform will coach the writer of an award message to be more comprehensive [and] more specific and nudge them to talk about the impact to the business that the person receiving recognition demonstrated. That helps the messages become richer, and the aggregated data will then be robust enough to serve highly accurate information when it’s queried.  

There are instances when the data is not high-value. One is when the recognition program does not have broad adoption. Say you’re only hitting half of the employee population. The other half is dark to the AI. It knows nothing about them. This goes back to program design, continuous communication about the program, and building motivation for people to use it regularly. 

Another example where the data would be lacking is when people write, “Bob did a great job.” There’s not a lot AI can do with that. That’s where the in-product coaching comes in to get into something more specific and descriptive.