4 ways to use AI in employee benefits, plus 3 drawbacks

4 use cases for AI in employee benefits administration

AI is most helpful for some specific aspects of employee benefits administration. Here are some use cases.

1. Answering employee questions

Implementing a chatbot to respond to workers' questions about their benefits can save time for HR employees.

Eric St-Jean

Using AI for benefits administration can make employees more self-sufficient because they won't have to reach out to HR staff with all their questions, said Eric St-Jean, an HR technology consultant at CleaHRStrat Consulting Inc., an HR consulting company located in Ottawa.

"[Questions about coverage] are things that AI can definitely be taught to answer without the employee having to always ask a person," St-Jean said.

2. Personalizing the benefits enrollment process

AI can also offer specific benefits options for employees based on their information.

AI can analyze premium costs, benefits usage predictions and other factors, then suggest a particular benefit plan for each worker, said Natalie Pierce, who works with organizations on how to incorporate generative AI and is a partner at Gunderson Dettmer, a law firm located in Redwood City, Calif.

AI personalizing their suggestions for workers can result in faster benefits processing times and a decreased workload for administrators, Pierce said.

The AI analysis of premium costs and benefits usage can also help employees save money.

AI might deduce that an employee with minimum to moderate healthcare benefits usage is best suited for a higher deductible, said Ron Hanscome, a research vice president at Gartner, a research and advisory firm headquartered in Stamford, Conn.

Employees will likely be happy to receive suggestions on how to reduce their benefits costs, resulting in a better employee experience.

3. Reducing benefits costs

AI's ability to offer insight into the popularity of certain benefits can also help companies save money.

Natalie Pierce

AI's predictive analytics can forecast healthcare utilization and employee benefits preferences, among other information, Pierce said.

For example, AI could share that very few employees used their employee lifestyle spending accounts in the last year. Company leaders could decide to cut a benefit if employees aren't using it.

"[AI] could allow employers to proactively adjust programs," Pierce said.

4. Providing better employee experience

In addition to improving employee experience by offering benefits suggestions, AI might help improve employee experience because of younger employees' potential expectations.

Younger professionals are expecting their employer's technology to help them make decisions based on potentially complicated information like benefits options, Pierce said.

"I think we're in a time where employees might wonder, 'Why don't we have virtual assistants?'" Pierce said. "'Why don't we have a chatbot that can help me figure out what is the best plan for me and my family?'"

3 risks of using AI for employee benefits administration

AI does come with potential issues, and HR leaders should make sure to consider them when implementing AI for employee benefits administration. Here are some potential problems.

1. Compromised data privacy and security

One of the risks of using AI for employee benefits administration is the possibility of employee data getting leaked through a cyberattack.

Privacy is another consideration for AI handling employee data, Pierce said.

Employee privacy must be a consideration from the beginning when developing AI to ensure that AI is only collecting and processing essential worker data.

In addition, HR and other company leaders should not simply trust that an AI vendor is taking care of security concerns.

Leaders must question their vendors about their approach to privacy and security, Pierce said. Vendors should submit data that confirms that they are carrying out what they say they are doing.

2. Poor integration with existing systems

While using AI for employee benefits could be helpful, the technology might not integrate with other company tools.

Ron Hanscome

However, a company's current vendors might add AI capabilities to their software in the future, avoiding the need for an organization to get separate AI tools.

HR leaders should ask their vendors about their near-term roadmap and whether they will be developing AI capabilities, Hanscome said. He believes waiting for a current HR software vendor to integrate AI features into its platform is better than investing in something new.

3. Dependency on AI

Some HR situations still require a human employee with empathy.

For example, employees might prefer to interact with a person during a discussion about leave management because the worker might be dealing with the loss of a family member or a medical emergency, Pierce said.

"Resource professionals can help employees navigate what might be very difficult times in their lives and their careers," Pierce said. "[That] would be an important time to keep the human touch."

In addition, AI technology is not perfect, and a human employee must check that AI is giving employees correct answers.

"If you're rolling out any type of AI functionality, there needs to be sufficient testing, at least initially," St-Jean said.

Conducting routine spot checks after that is also advisable to ensure the technology isn't making mistakes later, St-Jean said.

HR leaders should consider whether their company can reap the same benefits from different technology.

"You might be able to [achieve company goals] just as easily with something else and not have to deal with the pitfalls of early [AI] adoption," Hanscome said.

Carolyn Heinze is a Paris-based freelance writer. She covers several technology and business areas, including HR software and sustainability.