What Squeri actually said
American Express CEO Steve Squeri told analysts the company sees widespread gaming of its rewards and intends to do something about it. The remarks came on the Q3 2026 earnings call, which followed results released October 23, 2026.
The opening came from KBW analyst Sanjay Sakhrani, who asked how customers might gamify rewards and steer spending toward the richest benefits, and how Amex would respond. Squeri's answer was unusually direct:
"We have levers within our model, and we've shown this with the Platinum Card — where we can get to spend thresholds that require those benefits to be made, which then sort of eliminates that ability to game the system."
"I think in terms of rewards maximization, we've got a lot of gaming today that goes on."
On AI: it "will allow us to manage it a little bit better as well by using AI who understand what people are doing — and then how do we modify our value proposition."
"You've seen us do this with [the] $75,000 limit in terms of the pass for the Plat[inum]... you'll see other controls that are put into the decision."
He also drew a line that matters for reading the rest of this: "We really don't have a breakage model. So we don't benefit when people don't use our rewards."
One caveat on sourcing. These quotes come from a machine-generated call transcript with visible errors — the $75,000 line is garbled mid-sentence, and the bracketed words above are reconstructions. The substance is consistent across multiple outlets that covered the call, but anyone quoting Squeri verbatim in print should check the recorded audio first.
Sources: TradingKey transcript · View from the Wing · Amex Q3 2026 results, SEC 8-K
The $75,000 tell
Squeri named a specific precedent, and that precedent tells you the shape of what's coming. In February 2023, Amex stopped letting Platinum cardholders bring guests into Centurion Lounges free of charge.
Here is how that control works:
Platinum and Business Platinum cardholders pay $50 per guest, for up to two guests.
Authorized users on the account don't pay the fee.
The fee is waived entirely if the account spends $75,000 or more in eligible purchases in a calendar year.
Read it as a design pattern rather than a lounge rule. Amex didn't remove the benefit and it didn't means-test the cardholder. It attached a spending gate, so the perk stays available to anyone willing to put enough volume on the card, and quietly disappears for anyone holding Platinum mainly to extract benefits worth more than the fee.
That is what Squeri means by levers. When he says "you'll see other controls," the Centurion model is the template he just pointed at: not a smaller benefit, but a benefit you have to earn through spend.
Source: AwardWallet — Centurion Lounge guest access
What's already in motion
The crackdown didn't start on the earnings call. Three things were already underway before Squeri said a word, and together they show the direction of travel.
Fine Hotels + Resorts gets a two-night minimum. From January 1, 2027, the $300 semiannual Platinum lodging credit requires a booking of at least two nights. One-night stays qualify today. Eligible Airbnb Luxe stays are being added under the same two-night rule. This is aimed squarely at the one-night mattress-run booking made purely to trigger a credit.
Travel advisors were locked out of the booking channel. As of September 15, agencies can no longer secure FHR, The Hotel Collection or Cruise Privileges benefits through RezSaver. Qualifying reservations must go through American Express Travel. Amex said only that it "regularly evaluate[s] how our proprietary Card benefits are delivered" and that its own channel is "the best way" to administer them consistently.
The clawback language has been there for years. Amex terms have long allowed it to withhold or reclaim a welcome bonus where it determines you "engaged in abuse, misuse, or gaming in connection with this offer," and to deny the bonus if you cancel or downgrade within 12 months of opening. In March 2024 Amex tightened this further on personal Delta cards, adding language targeting applicants with a history of closing or downgrading cards inside a year.
The pattern is consistent. Amex isn't cutting headline benefits — it's closing the specific mechanics that let people extract maximum value for minimum commitment.
Sources: InsideFlyer · Forbes · Thrifty Traveler
What "other controls" probably look like
Squeri didn't specify, so everything in this section is inference, not reporting. Treat it as a way to think about risk, not a forecast — Amex has announced none of it.
That said, the template plus the existing friction points narrow the field considerably. The likeliest moves:
Spend gates on credits. The Centurion model applied to monthly and semiannual statement credits, where the richest ones unlock only above an annual spending threshold on the account.
Minimums and eligibility rules on credits. The FHR two-night rule is the first of this type. Dining, travel and retail credits with minimum transaction sizes, or exclusions for obvious gift-card workarounds, follow the same logic.
Tighter authorized-user economics. AU cards have been a well-known way to multiply benefits against a single fee. Per-AU fees and per-account rather than per-card benefit caps would close that.
More aggressive welcome-offer screening. The "pop-up" that tells applicants they're ineligible is already opaque and already driven by account history. A model with better behavioral data gets better at catching churn patterns before the bonus posts, not after.
Benefit caps that follow the account, not the card. The simplest structural answer to someone holding several premium cards for overlapping credits.
The honest summary: Amex told investors it has data, intent and precedent. It did not tell them — or us — what it will actually ship.
Source: Military Money Manual — the Amex pop-up
What this is really about
Follow the cost line. Amex rebuilt the Platinum card around a much larger stack of statement credits and took the consumer annual fee from $695 to $895 — a bet that cardholders would value the credits more than the fee increase.
The bet works only if redemption behavior stays within modeled ranges. Squeri's own framing gives the game away: with no breakage model, every credit a cardholder actually uses is a real cost. A population that uses every credit, every time, at minimum qualifying spend, is a population that costs more than the pricing assumed.
The Q3 numbers show a company with room to act but reasons to watch expenses:
Total revenues of $20.06 billion, up 9% year over year.
Net income of $3.71 billion, up 28%.
A pretax restructuring charge of $336 million tied to efficiency initiatives.
A roughly 5% workforce reduction, about 3,500 roles, announced alongside results.
CFO Christophe Le Caillec told analysts value-proposition spending should moderate once the Platinum refresh laps.
That last line is the one to watch. "Moderate once we lap the refresh" is a forecast about benefit costs, and the controls Squeri described are how a company makes that forecast come true if cardholder behavior doesn't cooperate.
The AI framing cuts both ways, which is why it came up at all. Analysts asked whether AI agents would help cardholders optimize harder; Squeri's answer was that the same technology lets Amex see what people are doing and reprice accordingly. Both things are true, and the issuer has far better data.
Sources: Amex Q3 2026 results · StockStory — Q3 analyst questions · NerdWallet — Platinum fee increase
What you can actually do
Nothing here requires panic, and most of it is good practice regardless. The goal is to look like a customer Amex wants rather than a cost line it is modeling.
Protect your welcome bonuses
Hold any card at least 13 full months before closing or downgrading. The 12-month clawback window is explicit in Amex's terms, and the safest exit is after the anniversary, within 30 days of the statement carrying the renewal fee.
Avoid closing several Amex cards in quick succession. A pattern of opening and closing inside a year is the specific behavior the 2024 Delta-card language was written to catch.
Check offer eligibility before you apply. If the application shows the pop-up saying you are not eligible for the welcome offer, abandon it — there is no credit-report consequence to stopping there.
Use benefits the way they are designed
Book Fine Hotels + Resorts stays of two nights or more from January 1, 2027. If you were holding a one-night FHR booking to trigger the credit, use it before year end.
Book FHR, The Hotel Collection and Cruise Privileges through American Express Travel. Since September 15, a travel advisor booking through RezSaver cannot deliver those benefits.
Spread credit usage naturally rather than running identical minimum-qualifying transactions on the last day of every month. Mechanical patterns are what behavioral models are built to flag.
Reduce your exposure
Re-underwrite each premium card on the benefits you would use even if the credits got harder to trigger. A card that only works when every credit is maximized has no margin for a rule change.
Avoid concentrating everything at one issuer. Chase, Citi and Capital One have their own enforcement, but a rule change at Amex should not be able to reprice your entire strategy.
Move Membership Rewards points you have a near-term use for. Points sitting in a program are exposed to that program's decisions; points transferred against a confirmed award are not.
Keep your own records. Screenshot offer terms when you apply and keep statements showing you met the spend. Clawback disputes are won with documentation.
One thing not to do
Do not preemptively close cards over this. Closing accounts inside 12 months is itself the trigger, so reacting to a crackdown by churning out of your cards creates the exact record you are trying to avoid.
What not to panic about
The coverage of these remarks has run hotter than the remarks deserve. Four things are worth keeping in view.
Amex announced nothing. Squeri described capability and intent on an investor call. No new control has been published, no date given, no benefit changed as a result of those comments.
"Gaming" is doing a lot of work. Squeri's own framing was about rewards maximization across cards — behavior most engaged cardholders consider ordinary. He did not define where optimization becomes abuse, and that ambiguity is the real story.
The no-breakage line cuts in your favor. A company that does not profit from unused rewards has little reason to make benefits harder to use in general. It has every reason to target the narrow behaviors that cost more than they return.
This is an industry pattern, not an Amex quirk. Chase reworked Sapphire economics this year, and Citi and American just repriced their co-brands. Issuers across the board are tightening after an expensive premium arms race.
The realistic read: engaged cardholders who use benefits on real trips are not the target. The exposure sits with people whose entire card math depends on extracting maximum credits at minimum qualifying spend.
Bottom line
The one move to make: re-underwrite every premium Amex card you hold on the benefits you would genuinely use if the credits got harder to trigger. If a card only pencils out when you squeeze every credit at minimum qualifying spend, that card is now carrying risk it was not carrying in September.
The one habit to drop: the mechanical minimum-spend credit run — the one-night FHR booking, the identical month-end transaction, the authorized-user card held purely to duplicate a perk. Those are the exact patterns Squeri said Amex can now see, and the two-night FHR rule shows it is already closing them one at a time.
Watch January 1, 2027. That is when the FHR minimum takes effect, and it will be the first real test of whether this is a one-off tightening or the opening move Squeri implied.
Izzy Hernandez, Founder, The Upgrade Life