What participants’ use of AI means for retirement plans

Artificial intelligence is already influencing how workers make financial decisions. The question for the retirement industry is no longer whether participants will use AI, but how advisors and plan sponsors can help them use it more effectively.

According to research from the Allianz Center for the Future of Retirement®, 70% of plan participants have used AI for financial advice or retirement planning, or would consider doing so.¹ More than half, 56%, expect their employer-sponsored retirement plan to incorporate AI capabilities within the next five years.¹

For advisors and plan sponsors, these findings point to a potential opportunity. AI could expand access to financial guidance, make retirement education more personalized, and help participants navigate decisions that often feel overwhelming.

But AI is not a substitute for financial literacy, thoughtful plan design, or professional support. In fact, participants’ growing use of AI may make those elements even more important.

Participants want guidance, but may not know what to ask

AI can explain financial concepts, organize information, and help users think through different choices. However, the quality of its guidance often depends on the context a participant provides.

That creates a challenge: People frequently turn to AI because they do not know how to approach a financial decision, but receiving useful guidance may require them to know which information matters and what questions to ask.

For example, retirement guidance can vary based on age, income, savings, debt, risk tolerance, taxes, health care needs, family circumstances, and retirement goals. If a participant leaves out important details, an AI tool may provide an answer that sounds personalized but is based on incomplete information.

The concern is amplified by gaps between how much participants believe they know and what they actually understand. While 78% of participants feel confident in their ability to make informed decisions about their employer-sponsored plan,¹ many hold fundamental misconceptions:

• 44% incorrectly believe their contributions are guaranteed to be available when they retire.¹
• 42% incorrectly believe they should hold more equities as they age.¹
• 35% incorrectly believe their investment returns are guaranteed to be positive and available at retirement.¹
• 35% do not know that delaying Social Security can result in a higher monthly benefit.¹

The implication is not that participants should avoid AI. It is that AI experiences should help participants build understanding, not simply deliver answers.

Younger participants may need a different approach

This issue is especially relevant for younger workers. Nearly 6 in 10 Gen Z participants have already used AI for financial advice.¹ At the same time, Gen Z participants recorded some of the highest rates of incorrect responses to questions about employer-sponsored retirement plans. Among Gen Z participants surveyed:

• 54% incorrectly believe their contributions are guaranteed to be available when they retire.¹
• 51% incorrectly believe they should hold more equities as they age.¹
• 51% incorrectly believe their investment returns are guaranteed to be positive and available at retirement.¹
• 49% do not know that delaying Social Security can result in a higher monthly benefit.¹

For advisors and plan sponsors, this presents both a risk and an engagement opportunity. Younger participants may be more willing to interact with an AI-powered tool than attend a traditional retirement seminar or read an educational brochure. Meeting them through a channel they already use could make retirement planning feel more accessible.

However, the experience should account for possible knowledge gaps. AI-powered education should define key terms, explain trade-offs, identify missing information, and encourage participants to seek additional support before making consequential decisions.

The goal should not be to make participants feel like financial experts after one interaction. It should be to help them make the next informed decision.

AI can help, but simplicity may come at a cost

Recent research from MIT suggests that AI-generated financial guidance can encourage positive behaviors, including saving more, establishing emergency savings, investing in diversified portfolios, and gradually reducing investment risk over time.²

That is encouraging, particularly for participants who have little access to individualized financial guidance. But the study also found that AI tends to rely on familiar rules of thumb.² As such, recommendations may default to round-number savings targets or commonly cited withdrawal approaches rather than fully accounting for an individual’s circumstances.

AI can also present generalized information in a confident, authoritative tone, making it difficult for participants to distinguish between a helpful starting point and a recommendation they should act on.

For plan sponsors, the takeaway is that AI should be considered part of the plan’s broader guidance framework. Its role, limitations, escalation points, and relationship to other advice services should all be clear.

For advisors, this creates an opportunity to help plan sponsors evaluate AI capabilities with the same discipline applied to other plan services. Questions to consider include:

• What participant data informs the experience?
• Does the tool explain why it provides a particular response?
• How does it identify missing or conflicting information?
• When does it direct participants to human support?
• How are its responses monitored for accuracy and consistency?

Why retirement income planning is an important test case for AI

Retirement income planning may be one of the clearest examples of both AI’s potential and its limitations. Approximately two-thirds of participants, 66%, say planning for retirement income feels overwhelming.¹ Participants must determine how to convert savings into income while considering longevity, market volatility, inflation, taxes, health care expenses, and Social Security.

AI may help explain these concepts, model different scenarios, and guide participants through the questions they should consider. It could also make retirement income education more timely by reaching participants as they approach retirement or begin evaluating distribution options.

But education and modeling do not solve the entire challenge. Participants still need access to products and plan features that support their goals.

More than half of participants say they would choose a lump-sum withdrawal if distribution or rollover options were unclear or required multiple steps.¹ 23% say they have no idea how they will manage their retirement savings – up from 18% in 2025.¹ These findings suggest that complexity can shape behavior, sometimes more than careful planning does.

AI cannot compensate for an experience that is difficult to navigate or a plan that offers few retirement income options. It can help participants understand an in-plan guaranteed lifetime income solution, for example, but AI-generated withdrawal guidance is not a substitute for the guarantees that an annuity can provide.

What advisors and plan sponsors can do now

The rise of AI does not necessarily require plan sponsors to implement a new tool immediately. It does require them to recognize that participants are already receiving financial information from sources outside the plan.

Advisors can help sponsors respond by focusing on four areas:

1. Evaluate the participant journey. Identify the decisions where participants are most likely to feel confused, leave the plan, or seek outside guidance.
2. Strengthen foundational education. Use plain language to explain investment risk, guarantees, Social Security, distributions, and retirement income.
3. Connect guidance to plan solutions. Ensure participants can move from learning about a financial need to understanding the tools available within their plan.
4. Preserve access to human support. Create clear pathways to education specialists, financial professionals, managed account advice, or other support when decisions become more complex.

The future is AI plus human guidance

Participants are not asking plans to make every decision for them. Only 17% prefer minimal involvement with their employer-sponsored plan, while most want moderate or high involvement.¹ They are looking for help that makes their involvement more manageable and meaningful.

AI can make guidance more accessible, personalized, and scalable. Human professionals can provide context, judgment, accountability, and support for decisions with lasting consequences.

For advisors and plan sponsors, a promising path is not choosing between the two. It is creating an experience in which AI helps participants ask better questions, human guidance helps them evaluate important choices, and thoughtful plan design gives them meaningful options on which to act.