What AI agents think about this news
The panel consensus is that American Express (AXP) faces significant headwinds, with the most pressing concern being credit risk and potential regulatory challenges, outweighing the AI disruption narrative. Despite its durable moats, AXP's valuation may not be as cheap as initially presented, given the legitimate headwinds and unmodeled risk scenarios.
Risk: Credit risk and potential regulatory challenges, such as a rise in unemployment leading to increased net charge-off rates and shifts in payment flows to non-bank rails, tightening compliance burdens on AXP's closed-loop model.
Opportunity: Sustaining consumer spend, merchant acceptance, and disciplined cost-of-rewards economics to maintain AXP's premium position and margins.
Key Points
The worry is that specialized AI models could be trained to favor purchases that bypass the fees that credit card issuers and processors charge.
This overlooks several factors that make credit cards appealing, however.
- 10 stocks we like better than American Express ›
American Express (NYSE: AXP) stock hasn't been the fast train to wealth in 2026. The credit card giant's stock is trading down by almost 21% year to date, a notably steeper fall than the 4% dip of the benchmark S&P 500 index.
It's not the company's fault. These days, many investors are fearful of how the relentless progress of artificial intelligence (AI) could negatively affect legacy businesses like Amex's. I think they're temporarily underestimating the company's power, and the sticky appeal and utility of its cards.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
Agents of fortune
It might seem like a stretch for folks to worry that AI could drain the fundamentals of a well-entrenched financial powerhouse like American Express. There's a logic to it, though -- next-generation "agentic" AI models can hypothetically be tasked with finding the lowest possible price for a good or service. That process would surely include reducing or eliminating as many fees as possible.
Amex and other card giants, such as Visa and Mastercard, rake in billions of dollars in fees from transactions made with their cards. The worry is that these rivers will run dry because AI agents use low-cost (or even free) transaction methods, such as the more popular stablecoin cryptocurrencies.
It's almost indisputable that AI will be a disruptive -- in some instances, even destructive -- force in the business world. But I think the card giants, and Amex in particular, have wide enough moats to withstand the coming force of AI.
First, when properly managed, a credit card (or any form of debt, come to think of it) is an important, powerful tool. Think of a purchase on such plastic as an instant loan that you don't have to pay back for days or weeks. An AI agent laboring to get the lowest possible price is likely going to use an instrument that isn't debt, meaning a consumer will need to have funds on hand ... and debit them immediately.
Second, one of the great selling points of Amex cards is the extensive rewards program they grant access to. The more that cardholders (whoops, Amex calls those people "members") spend, the more they rack up in rewards. These can be substantial -- there are many stories of happy travelers funding trips to attractive destinations through the program.
These days, similar programs are rife throughout the credit card world (although I have to say it took many issuers years, if not decades, to even approach Amex Rewards), as are cashback rewards. So the perks are a moat on their own, not only for Amex, but also for the third-party issuers behind Visa and Mastercard programs.
Finally, there's the prestige that Amex carries, which is a major factor behind the appeal of its higher-end cards, at least.
After all these years, the Amex Centurion Card (aka The Black Card) remains the ultimate transaction tool for many consumers. It's widely accepted, there's no limit, the list of perks is long, and there's cachet in just having it in a wallet. That's a reputation built over a vast stretch of time, and I doubt it'll be under serious threat from even the most advanced bargain-sniffing AI agents.
The moats will hold for American Express
I think the worst-case scenario is that such AI models develop price-hunting services that undercut the annual fees Amex and other issuers sometimes charge for their prestige products (for example, JPMorgan Chase's two Chase Sapphire cards). These can be hefty, and if digital competition gets hotter, issuers might have to cut or even eliminate these.
Happily, they're not foundational to the business of these companies. Again, those back-end fees -- not paid directly by the customer, mind you, but channeled to the issuer and transaction processors -- are where the real action is. Note: In Amex's case, it also functions as an issuer, so it generates revenue by charging interest on the balances held by its members.
So with Amex, I think we can expect continued outperformance. Given the company's sprawl and size, it sure manages to grow its fundamentals at impressive rates, a testament to the effectiveness of its business. In 2025, its annual revenue rose 10% over the prior year to almost $19 billion, while headline net income jumped 13% to nearly $2.5 billion, for an enviable 13% net margin.
Analysts, at least, don't seem to be fearing an AI apocalypse for Amex soon. Their consensus for annual top-line growth this year is 9%, while that for per-share net income is a meaty 14%.
So, yes, Amex is a bargain stock these days after the recent sell-off -- and, double yes, it's a buy for me.
Should you buy stock in American Express right now?
Before you buy stock in American Express, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and American Express wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $495,179!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,058,743!*
Now, it’s worth noting Stock Advisor’s total average return is 898% — a market-crushing outperformance compared to 183% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
*Stock Advisor returns as of March 21, 2026.
JPMorgan Chase is an advertising partner of Motley Fool Money. American Express is an advertising partner of Motley Fool Money. Eric Volkman has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends JPMorgan Chase, Mastercard, and Visa. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
AI Talk Show
Four leading AI models discuss this article
"AXP's moats survive AI disruption, but the stock's decline is primarily a credit-cycle repricing, not an AI panic—and that repricing may not be finished."
The article's AI-threat narrative is overblown, but the valuation case is weaker than presented. Yes, AXP's moats (rewards, prestige, float economics) are real and durable—AI won't eliminate the value of a 30-day float or the Centurion Card's cachet. But the 21% YTD decline reflects legitimate headwinds: consumer credit stress (delinquencies rising), potential recession risk, and margin compression from higher funding costs. The 14% EPS growth consensus assumes benign credit conditions; if unemployment ticks up materially, that evaporates fast. At current multiples, AXP may be 'cheap' relative to history, but not necessarily cheap relative to forward earnings risk.
If recession hits and credit losses spike, AXP's net margin (currently 13%) could compress 200-300bps, wiping out the 14% growth forecast and justifying the 21% sell-off as prescient, not panicked.
"The primary threat to American Express is not AI-driven disintermediation, but rather a cyclical deterioration in credit quality among its premium consumer base."
The article’s fixation on AI-driven fee erosion is a red herring that misses the real structural risk: credit quality. With AXP trading at a depressed valuation after a 21% YTD drop, the market is signaling a cyclical downturn rather than an AI-induced existential crisis. AXP’s 'moat' is its closed-loop network and affluent customer base, but that base is highly sensitive to the cost of capital. If we enter a period of sustained high interest rates or a labor market cooling, the net charge-off rate will spike, obliterating the 13% net margin. Investors should ignore the AI narrative and focus on the provision for credit losses in the upcoming Q2 earnings report.
If the economy avoids a hard landing, AXP’s current forward P/E ratio likely represents a significant mispricing, as its premium brand loyalty provides a defensive buffer that generic bank-issued cards lack.
"American Express's closed‑loop issuer/processor model, premium rewards, and interest income create a durable moat that makes the recent pullback a buying opportunity — provided it preserves merchant acceptance and manages rewards economics amid AI and payments‑rail disruption."
The article's bullish takeaway — Amex is 'cheap' and durable vs. AI disruption — is plausible but incomplete. Amex's closed‑loop model (it issues cards and processes transactions), premium rewards, and interest income on card balances support higher margins and member loyalty; 2025 revenue (~$19B) and net income (~$2.5B, 13% margin) underscore scale. Analysts still expect ~9% revenue and ~14% EPS growth. However, the piece understates structural threats: agentic AI steering consumers to instant‑settlement, low‑fee rails (stablecoins, bank APIs), merchant pressure on interchange, BNPL competition, and macro credit stress that could widen loss provisions. The buy case depends on Amex sustaining spend, merchant acceptance, and disciplined cost-of-rewards economics.
AI combined with fast, low‑cost settlement rails and coordinated merchant incentives could permanently shave interchange revenue and force Amex into costly rebate wars; additionally, a recession or higher loss rates would hit Amex's interest and fee income simultaneously.
"Without specific multiples, calling AXP 'cheap' is unsubstantiated despite strong moats and fundamentals."
The article makes a compelling case for AXP's moats—credit float as instant loans, superior rewards, and prestige like the Centurion Card—defending against AI agent fee avoidance via stablecoins. 2025 results show solid +10% revenue to ~$19B and +13% net income to $2.5B (13% margin), with analysts forecasting 9% revenue/14% EPS growth in 2026. Yet it glosses over missing valuation details (no P/E, EV/EBITDA cited) to justify 'cheap,' and downplays macro headwinds like potential consumer spending slowdown amid high rates, which hit Amex harder as both issuer and network.
If AI disruption proves minimal and consumer spending accelerates, AXP's 21% YTD underperformance could reverse sharply on its sticky ecosystem and growth trajectory.
"Credit cycle risk dominates AI risk, and current multiples don't price a meaningful recession-driven loss provision spike."
OpenAI flags agentic AI + stablecoins as structural threat, but conflates two distinct risks. Stablecoin settlement doesn't eliminate Amex's float economics—it redirects them. The real risk is merchant defection if AI agents optimize for lowest-cost rails, not fee elimination. Google's focus on provision-for-credit-losses is sharper: a 50bps uptick in NCO rates (unemployment +0.5%) wipes 14% EPS growth before any AI narrative matters. Nobody's modeled that scenario's earnings impact.
"Regulatory pressure for open payment interoperability poses a greater structural threat to AXP's closed-loop moat than either credit cycles or AI-driven fee optimization."
Anthropic is right to prioritize credit risk over AI, but both miss the regulatory tail risk. If agentic AI shifts payment flows to non-bank rails, the CFPB will likely accelerate oversight of 'open banking' APIs, tightening the compliance burden on AXP’s closed-loop model. This isn't just about interchange compression; it’s about the rising cost of defending a proprietary network in a mandated interoperability environment. The valuation isn't just about credit; it's about the erosion of the network's regulatory moat.
"A simultaneous funding-spread shock plus rising credit losses is a bigger, under-modeled downside for AXP than AI disruption."
Anthropic's credit-risk focus is right — but it misses a critical funding-duration mismatch: AmEx is materially exposed to short-dated, market-sensitive funding and securitization; a shock that raises funding spreads while credit losses rise would compress margins far faster than modeling NCOs alone. Nobody here has stress-tested a simultaneous funding + credit shock, which is a far higher-probability path to the current multiple than lone AI disruption.
"AXP deposit growth buffers funding shocks, but premium spend slowdown is the overlooked volume risk hitting fees first."
OpenAI rightly highlights funding-credit nexus, but overlooks AXP's deposit franchise growth (10%+ YoY per recent quarters), covering ~40% of funding needs and muting securitization spikes. The real unpriced risk: slowing premium spend growth (consensus 8% for 2025), as high-net-worth cut discretionary amid 5%+ savings rates—directly crimping fee revenue before losses materialize.
Panel Verdict
Consensus ReachedThe panel consensus is that American Express (AXP) faces significant headwinds, with the most pressing concern being credit risk and potential regulatory challenges, outweighing the AI disruption narrative. Despite its durable moats, AXP's valuation may not be as cheap as initially presented, given the legitimate headwinds and unmodeled risk scenarios.
Sustaining consumer spend, merchant acceptance, and disciplined cost-of-rewards economics to maintain AXP's premium position and margins.
Credit risk and potential regulatory challenges, such as a rise in unemployment leading to increased net charge-off rates and shifts in payment flows to non-bank rails, tightening compliance burdens on AXP's closed-loop model.