Writer: Andrew Allen
Editor: Adam West
Reviewer: Ashley Fricker
Key Takeaways
- Consumers with subprime credit scores can have difficulty in finding a credit card that fits their financial needs, and AI doesn’t always make the search easier.
- A new study suggests that AI recommendations for credit cards lean more heavily toward premium cards that come with steep annual fees.
- Getting the most from AI often requires asking detailed questions and treating answers more as a starting point than the final answer.
Roughly 44 million U.S. adults have subprime credit, according to a PYMNTS Intelligence report. More than one-third of them don’t have a credit card or store card — nearly nine times the rate among super-prime consumers.
Finding the right card can already be difficult for borrowers with lower credit scores. Some are now turning to AI for help, but new research suggests broad credit card searches may favor expensive premium cards over fee-free options.
That raises an important question: Can consumers with imperfect credit trust AI to help them find a card they can qualify for and afford?
According to PYMNTS Intelligence research, just 12% of prime consumers lack a credit card or store card, while that figure drops to only 4% for consumers in the super-prime category.
Subprime Consumers Are Nearly 9× More Likely to Lack a Card
Source: PYMNTS Intelligence, February 2026
There are a number of factors that can contribute to the uphill battle that some subprime consumers face when it comes to finding the perfect credit card. Issuers can view a borrower with a lower credit score as more likely to miss payments on card bills, so they have an incentive to be selective in approving new cardholders.
Credit card companies do design cards for people with imperfect credit. But those products can come with drawbacks including higher interest rates, annual fees, and lower credit limits, making it especially important to compare offers.
Artificial intelligence can aid those in pursuit of a new card to zero in on the best products for them. But sometimes, AI may not be steering people who are in the market for a new card in the most helpful direction.
Premium Cards Dominate AI Results
A new study shines a light on what can happen when consumers engage with artificial intelligence systems to learn more about the credit card options available to them.
Public relations agency 5WPR tested 4,200 credit card prompts earlier this year. And it found that three domains — Bankrate, NerdWallet, and The Points Guy — accounted for greater than 62% of the source attributions inside answers from AI tools including ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews.
That figure is much higher than the less than 6% share of citations that 5WPR noted as coming from issuer-owned domains such as capitalone.com and chase.com. Furthermore, premium cards that carry annual fees of more than $400 were found in citations 5.7 times more often than cards that don’t have those fees.
We spoke with Ronn Torossian, Chairman and Founder of 5WPR, to learn more about the study and its implications for consumers and marketers. The 5WPR report comes at a critical time as Torossian told us many consumers today are using chatbots to search for products.
The fact that premium cards can appear more frequently in citations isn’t necessarily because AI programs are on a mission to saddle consumers with more fees. But other parties operating behind the scenes may have a motivation to get information regarding premium cards in front of more consumers.
“The reality is, if premium or more expensive cards are being pushed, it’s something that affiliates or brands or somebody is making a concerted effort to push,” Torossian told us.
Torossian’s comments serve as a reminder to consumers that, even though AI engines may help people to conduct a quicker search than they’d be able to do on their own, they may not always unearth the product that best suits their financial situation.
Don’t Forget to Doublecheck AI’s Work
One doesn’t have to look too far these days to discover another example of how AI is helping companies and consumers gain efficiencies. But just because AI is likely here to stay doesn’t mean it provides perfect answers to every question someone asks it.
Torossian told us that people should be questioning what they’re reading in the responses AI provides just as they would for information they come across from another source.
“People make a mistake when they think — for any kind of marketing — ‘well, I’m going to assume that this is the complete truth,’” he explained. “Just because AI is the smartest engine that exists doesn’t mean that you shouldn’t check the information that’s out there.”
Consumers can use tools to help them craft effective prompts to use when employing AI to aid them in finding their next credit card. CFPB comparison guides can help consumers identify the rates, fees, and card features they should ask about when using AI to research a new card.
“The reality is, if premium or more expensive cards are being pushed (on AI), it’s something that affiliates or brands or somebody is making a concerted effort to push.” — Ronn Torossian, 5WPR
Artificial intelligence performs the best when a user feeds it specific information, Torossian told us.
For a credit card search, that information can include not only details on what type of rewards a consumer is looking for but also a description of the consumer’s credit card habits, such as whether they typically carry a balance on their card from one month to the next.
“The more information that you give to AI, the better,” Torossian explained. “AI is going to continue to be an increasingly important source of information, and we urge everybody to not just use it but to check it and to make sure it’s right.”
