Reenie Hankerson lives alone in the Bronx, N.Y., but the 69-year-old retiree never lacks for companionship.

Hankerson shares her apartment with ElliQ, an artificial-intelligence device inspired by the lamp in Pixar's logo. A history buff, Hankerson sometimes talks to ElliQ about the civil rights movement or women's suffrage. She plays games through ElliQ as well, including live bingo with other people around the country.

"It's good to have someone to talk to. She's a someone, even though she's a robot," says Hankerson, a retired crossing guard and home health aide.

As AI ripples through the economy and society, it's starting to play a role in retirement. Companionship has emerged as one of the more promising apps, but it's only one way AI is reshaping the way Americans plan for retirement and handle two of its most important facets: healthcare and investing.

The technology is rapidly getting smarter. AI-infused chatbots have memorized medical knowledge from the internet and can draw connections to individuals' medical histories and symptoms. Behind the scenes, hospitals are using AI to help monitor patients and keep them safe. In assisted living facilities, AI shows promise in helping prevent falls, a leading cause of frailty and death among older Americans.

Investors have a lot riding on AI. By market cap, nearly half of the U.S. equity market is in AI stocks. The corporate bond market is getting an AI infusion as companies like Alphabet, Meta Platforms, and Oracle issue billions in debt to fund their growth. Even if you don't directly own tech bonds, they're showing up in funds and could impact the broader market's performance.

As the technology expands, it's important to understand its limits-and how it cuts both ways. Insurance companies, for instance, are using AI to process claims and help determine what they'll cover, a practice the government is also testing in Medicare. The software may curb frivolous billing, but it could yield denials of legitimate claims. And if the AI stock party fizzles? A selloff could cascade and drag down the broader market.

Here's a look at some ways AI is transforming retirement and what you should know.

How AI Is Changing Healthcare

Chatbots like Google's Gemini or OpenAI's ChatGPT can be great tools for basic medical information. They can remember your conversations, prompts, and medical history, and draw connections to your symptoms. The underlying intelligence has, in some instances, diagnosed a rare disease that doctors missed.

One clear benefit is distilling medical jargon into plain English. For a loved one with moderate cognitive impairment, AI can provide a second-grade-level synopsis of a doctor's report, for example.

"They're incredibly good at that," says Arjun Manrai, associate professor of biomedical informatics at Harvard Medical School. Another useful application is enlisting the bot to help you prepare for a doctor's appointment by brainstorming questions to ask.

"The chatbots provide a judgement-free way to ask questions," says Carri Chan, professor of healthcare management at Columbia Business School.

Yet like humans, the bots are prone to error. And AIs themselves warn they pose "significant technical limits and serious privacy trade-offs," as Anthropic's Claude put it.

As the bots play doctor, lawsuits have begun. In one recent case, a Florida pastor sued OpenAI, alleging ChatGPT minimized symptoms that ended up being a life-threatening pulmonary embolism. An OpenAI spokesperson didn't comment on the lawsuit but noted in a statement to Barron's that, "ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis, or treatment."

Take a commonsense approach, experts say. Use the chatbots for basic information, but don't outsource your care to Dr. AI. "Whenever the stakes are high, you have to get a human involved," says Dr. Jonathan Chen, director for medical education in artificial intelligence at Stanford Medicine.

Less visibly, AI is operating behind the scenes in doctors' offices and hospitals. You might notice your doctor making more eye contact because an AI is listening to the conversation and taking notes. In some hospitals, systems can monitor vital signs to gauge patients' risk of winding up in the intensive care unit.

"It's a crystal ball for the next 24 hours," says Dr. Lee Schwamm, chief digital health officer at the Yale New Haven Health System, which is using the technology. Typically, some 5% to 10% of its inpatient population will deteriorate within 24 hours and end up in the ICU, Schwamm says. Identifying their risk earlier with AI can lead to preventive measures.

Ideally, AI will curb healthcare cost inflation without compromising care. But the technology may now be raising costs, analysts say, as hospitals like Yale New Haven invest while waiting for a payoff that doesn't come at patients' expense. At the same time, AI-enabled tools are helping medical providers capture more revenue, contributing to an estimated 9% increase in health costs next year, according to a PwC report.

Insurers, meanwhile, are using AI to help decide whether to approve a service or require prior approval before a procedure. Medicare Advantage insurers such as UnitedHealth and CVS Health are facing bipartisan calls in Congress for more information amid concerns the technology will be used to increase claim denials and delays.

The federal government has also begun using AI to help determine coverage in traditional Medicare. Under a pilot program that began on Jan. 1, prior authorization is required for a handful of procedures in six states. The AI-powered initiative, called the Wasteful and Inappropriate Service Reduction, or Wiser, Model, aims to cut costs for taxpayers. The government has said that "human clinical review" is part of its AI program, which will help protect taxpayers from wasteful spending.

Patients who feel their coverage has been improperly denied can appeal. A government report that looked at admissions to skilled nursing facilities under Medicare Advantage found that few patients appealed their rejection but among those who did, an overwhelming number won their petition.

An AI Infusion In Caregiving

If you're looking for a frilly robot like the Jetsons' Rosie, you'll have a long wait. But AI-powered caregiving and robotics are showing progress.

One way it's happening: helping to prevent falls. At Aegis Living, a network of 37 assisted living facilities in the Western U.S., an AI system resulted in 22% fewer falls during a 12-month period and 24% fewer injuries due to falls, according to Aegis founder and CEO Dwayne Clark. Aegis' residents lived 63 days longer, on average, Clark told Barron's.

"When we got the data, I was shocked," he says.

It will probably take more than a decade for an AI-powered robot to help older adults with all activities essential for daily living, from getting out of bed to bathing, says Steve Cousins, executive director of the Stanford Robotics Center.

But what about a robotic dog? It's in the works. Cousins' center is building prototypes for a low-maintenance robotic dog that seniors could walk around the neighborhood. "If you keep moving, you stay independent," Cousins says.

The humble wheelchair could also get an AI upgrade. ATDev, an assistive robotics firm, is working on an AI-powered wheelchair device that can do things like fetch water for users. "Devices are about giving agency back to people," says ATDev co-founder Owen Kent, who lives with muscular dystrophy and uses a wheelchair.

AI may also integrate devices in your home so they can talk to each other, says Andy Miller, senior vice president of innovation and product development at AARP. "The power of AI isn't in individual solutions; it's in the connectivity."

Sensors could detect that a person's balance is off, for example, and tie that to a missed dosage of a medication. The system might then issue an alert or send a report to an emergency contact.

The key, consumer advocates say, will be to bolster seniors' autonomy instead of eroding it. There's risk that AI technologies could replace human caregivers in nursing homes and other care facilities. "This is a technology with no guardrails around it," says Alex Lawson, executive director of Social Security Works, an advocacy group. Regulations to curb AI's potential for damage in healthcare will be needed, he adds.

How to Invest in AI-and Stay Safe

Retirees face a conundrum with AI: You could pay a steep opportunity cost if you avoid it and the tech sector keeps rallying. Conversely, even if you don't own tech stocks or funds, you could incur heavy losses if tech earnings forecasts wane, dragging down the broader market.

Tech could also sell off because of another growing risk: rising bond yields. Globally, yields are inching up as investors worry about unsustainable debt levels and inflationary pressures. Rising yields exert pressure on stock prices. U.S. markets may be at a tipping point, hinging on the Federal Reserve's next moves with interest rates and monetary policy.

Consider tailoring your AI exposure to your age and personal situation. Investors in their 20s, 30s, and 40s have plenty of time to weather a bursting AI bubble. A model portfolio for a 20-something could hold as much as 90% in stocks, advisors say, including funds tracking the S&P 500 or other tech-heavy indexes. The rest could be in cash and short-term bonds, the latter looking relatively secure.

Investors in their 50s, 60s, and beyond could take more balanced approaches-owning 60% to 40% in stocks, with the rest in bonds and other shock absorbers to help preserve capital.