Anthropic Tops US Business AI Adoption for the First Time — Ramp AI Index, June, Overtakes OpenAI
In Ramp's June AI Index — built on real corporate-card spending, not surveys — Anthropic hit 41% and passed OpenAI (32-33%) for the first time. The gap has widened since the first flip in April, and ~70% of companies buying AI for the first time skip OpenAI and go straight to Claude. For an IPO-bound Anthropic, that's a killer card.

It Finally Happened — US Businesses Are Swiping Their Cards for Claude More Than OpenAI
Here's the deal: in June 2026, among US companies paying for AI subscriptions, Anthropic hit 41% and edged past OpenAI (32-33%) for the very first time. That's the whole headline. One sentence. The catch is that this one sentence is a lot heavier than it looks.
Why heavy? Because this isn't a survey. Nobody asked "which AI do you use?" and counted raised hands. This is real money — actual corporate cards swiped and invoices paid by real companies. Plenty of firms say "oh, we're an OpenAI shop" while quietly spending their budget on Claude. Spending data strips away that little bit of hypocrisy and just shows you the money. That's exactly why this number is scary, and why the whole industry collectively went huh.
And this isn't a "led for one month, no big deal" thing either. The first flip showed up in April (Anthropic 34.4% vs OpenAI 32.3%), and since then the gap hasn't shrunk — it's gotten wider. This isn't a one-off blip; it's a trend hardening into place. The same OpenAI whose ChatGPT brand conquered the entire consumer market is starting to lose on the one battlefield where the money is actually serious: B2B.
So here's the story we're going to unpack today. How a perennial runner-up flipped the enterprise market in a single year, what OpenAI was doing the whole time, and why — with an IPO right around the corner — this number is the best card Anthropic could possibly be holding. Follow a few numbers and a few characters and the whole picture snaps into focus.
The Cast — Anthropic, OpenAI, and the Yardstick Called the Ramp AI Index
First, Anthropic. For a long stretch they lived in ChatGPT's shadow with a "safe but kinda quiet" reputation. Claude was clearly smart, but on public name recognition it wasn't even a contest against ChatGPT. Except they'd been quietly digging in somewhere else entirely: businesses, and specifically developers and coding. As Claude Code caught fire among engineers, a vibe formed — "the public may not know it, but the people who actually do the work all run Claude."
Next, OpenAI. Yeah, that OpenAI. No introduction needed — the company that jammed AI into everyone's daily life via ChatGPT, the undisputed king of the consumer market. They still dominate among regular users. But that turned out to be a trap. A consumer brand that strong can actually obscure the "serious B2B tool" identity. And crucially, over the past year OpenAI's business adoption barely budged. We'll see it in the numbers below, but that's the real shock.
Finally, today's actual protagonist: the Ramp AI Index. Ramp is a corporate-card and spend-management company, which means they get to watch what a huge number of businesses actually pay for. They roll that spending data up into the Ramp AI Index. The whole point is that it's real payments, not opinions. Behavior, not talk. Money, not words. That's why the market treats it as one of the most credible signals out there for "which AI products are businesses actually opening their wallets for."
Put the three together in one line: an objective money-tracking yardstick (Ramp) officially confirmed that the perennial runner-up (Anthropic) has overtaken the undisputed king (OpenAI) in the enterprise market. That's the backbone of today's story. Now let's look at the numbers that yardstick spat out.
The Core — The Flip, in Numbers
Talking forever beats nothing, but one table beats talking. Here are the key metrics of this reversal.
| Metric | Anthropic | OpenAI |
|---|---|---|
| June business adoption (paid subs) | 41% | ~32-33% |
| Adoption at the April first-flip | 34.4% | 32.3% |
| Past-year business adoption growth | ~4x (quadrupled) | ~+0.3 pts (basically flat) |
| First-time AI buyers going direct | ~70% (head-to-head wins) | — |
Let's rip through it row by row. First, 41% vs 32-33% in June. That's a gap of roughly 8-9 points. In April the gap was 2 points, so it's quadrupled in two months. The first message of this table: "Anthropic didn't just pull ahead once — it keeps pulling further ahead." The trend has tilted Anthropic's way.
The genuinely scary row is the growth line. Anthropic quadrupled its business adoption in a year. OpenAI, over the same stretch, grew just 0.3 points. Meaning OpenAI didn't shrink — it just barely grew at all in the enterprise market. In that window, Anthropic scooped up nearly all the new demand. Share fights ultimately come down to "who eats the incoming volume," and Anthropic all but monopolized it.
Third row, 70% going direct. Personally I think this is the most symbolic one. Among companies that had never adopted AI before — the so-called "never-adopters" — about 70% pick Anthropic when they compare the two for their first purchase. They're not trying ChatGPT first and migrating; they're skipping OpenAI entirely and going straight to Claude. That's a sign the default for new customers is shifting, which is why it makes you weight the long-term trend more heavily than near-term share.
Who Gains What
The most direct winner is obviously Anthropic. But this isn't a "we're number one, yay" moment. Anthropic is prepping an IPO and has already filed a confidential S-1. What does a company about to go public need most? Objective proof that "we're winning where the money actually is." Not a survey, not a self-published press release — a third party (Ramp) showing them at number one via real payment data. As ammunition for convincing investors, it doesn't get much better than that.
And this gain compounds because it plugs into a story. OpenAI owns consumer, but we own enterprise → businesses spend consistently and churn less → so our revenue is sturdier and more predictable. That logic chain bolts right onto the IPO valuation narrative. The developer and coding turf that Claude Code represents is especially sticky — once it's wired into a workflow it's hard to rip out — which lends an impression of "high-quality, lock-in-heavy revenue."
On the flip side, what OpenAI loses is less the share itself and more the narrative. OpenAI used to just be the synonym for AI, period. Now that a "consumer = OpenAI, enterprise = Anthropic" framing exists, the invincibility halo cracks a little. To be clear, the consumer market is so massive that this is nowhere near a death sentence. But shifting from "number one everywhere" to "a giant that got overtaken in one arena" stings in a field this competitive.
There's also a surprise beneficiary: Ramp itself. This episode firmly established the Ramp AI Index as "the AI-adoption barometer the industry watches." A company that owns spending data turning that data into an industry benchmark to grow its influence is a textbook move. And for observers like us, it's not a bad thing either — we just got one more indicator that watches money instead of words.
Echoes From History — Wins and Losses
This picture — an underdog flipping the leader in the enterprise market — isn't new. It's a pattern tech history has run again and again. The classic success story is the cloud market. People once figured "enterprise IT goes to the giant integrators and incumbents in the end," but a service that developers found easy to start using clawed its way up from the bottom and reshaped the entire market. "If developers adopt it first, the company budget follows" — Claude Code is walking exactly that path right now.
Another one that comes to mind is the browser and collaboration-tool wars. At first there was a high-recognition incumbent, then the practitioners quietly switched to whatever made the work go better, and that tool became the standard. The core lesson is identical: public recognition and on-the-job adoption are two different games. However famous ChatGPT is with the general public, if an engineer feels Claude is better for writing code, the payment goes to Claude. That's precisely what this Ramp data showed.
But to be fair, you have to stare just as hard at the failure cases. Plenty of products hit number one in share one quarter and got bumped right back the next. The software market has relatively low barriers and fast-shifting switching costs, so "number one this month" doesn't guarantee "number one forever." AI especially is a place where model performance sloshes around on a monthly cadence, so a rival dropping one better model can flip the mood all over again.
So the balanced read on this reversal is this: the direction is real and the meaning is big, but it's not a stamped, permanent win. The trend is clearly tilted toward Anthropic — that's true. But for that trend to hold six months out, Anthropic has to keep winning, and that's an unproven future. The lesson the past cases hand us is exactly this sense of balance.
The Competitor's Counter-Play
So is OpenAI just going to sit there? Of course not — they have cards to play. The first is dragging consumer power down into the enterprise. ChatGPT is so familiar to regular people that OpenAI can amplify the "let's just use what we already use" pull from employees inside companies. That's bottom-up infiltration, where individual users tug company adoption along, and OpenAI's weapon here is far stronger than Anthropic's. Consumer recognition is essentially a built-in enterprise sales pipeline.
The second is going straight at the coding and developer turf. The core engine behind Anthropic's surge right now is Claude Code. OpenAI will pour firepower into coding tools and the developer ecosystem to reclaim that territory. The developer market is sticky once a workflow sets in, but flip that around and it's also a market where a clearly better tool gives people a real reason to switch — so the match isn't over.
The third is price, bundle, and integration pressure. Enterprise customers are ultimately sensitive to "performance per dollar" and "fit with the tools we already use." If OpenAI cuts prices aggressively or wedges itself deeper into a massive distribution channel like the Microsoft ecosystem, companies have fewer reasons to bother switching. Since the adoption fight isn't a pure model-performance game but a sales, distribution, and ecosystem game, OpenAI's sheer size is still a scary weapon.
And don't forget the variable: the third players. Google and other big tech, plus open-source and China-origin models, are all eyeing the enterprise market. It looks like an Anthropic-vs-OpenAI two-horse race now, but if companies lean into a multi-model strategy — mixing several AIs by situation — the meaning of "the single number one" itself fades. So this reversal is best read not as the end of the game, but as the starting whistle for the real scramble over the enterprise market.
So What Actually Changes — By Who You Are
If you're an investor. The key is that Anthropic's IPO story just got some objective meat on it. Third-party payment data confirmed enterprise number one, so it's easier to bake a "business revenue is sturdy" logic into the valuation. Don't get carried away, though. Number one this month isn't number one forever, and OpenAI's consumer revenue is still enormous — don't forget that. The trend is good, but whether it's hardened into lock-in takes a few more quarters to confirm.
If you're an enterprise decision-maker. The "everyone uses ChatGPT so we should too" era should be considered over. The money from companies that actually do the work is increasingly flowing to Claude, especially in coding and developer workflows. That doesn't mean switch blindly. The point is choose based on your actual work, not brand recognition. If coding is a big chunk of what you do, there's plenty of reason to seriously test Claude — and where possible, don't bet it all on one provider; staying flexible with multiple models is better for risk management.
If you're a general observer. The real fun of this episode is that it shows the AI market isn't one market — it's several. Consumer and enterprise can have different winners, and right now they actually do. Going forward, when you read AI news, asking "number one in which market, by what measure?" instead of "who's most famous?" gets you a much sharper read on the board. You probably also felt firsthand this time why an indicator that watches money instead of surveys is so powerful.
The one line cutting across all three: the direction is clear, but the conclusion is still open. Anthropic breaking out in the enterprise market is real and meaningful, but whether that's set concrete or drying paint is something the next batch of Ramp data will tell us. So rather than overreacting, calmly watching the trend over the next few months is the smartest posture.
🥄 Three Things You're Probably Wondering
— So did Anthropic actually beat OpenAI? In one arena — business adoption, and specifically by the "spending data Ramp can see" — yes, clearly. A 41% to 32-33% gap leaves no real doubt. But saying "they won the AI war" would be an overstatement. OpenAI still dominates consumer. It's accurate to say they won a battlefield, not that they ended the war.
— Could this flip back next month? Honestly, it could. Software markets switch fast, and AI is a place where one better model swings the mood. That said, the gap actually widening after the April first-flip looks more like a trend than a coincidence. Still, it's too early to declare "this trend holds six months from now." We'd need to see a few more Ramp data points to be sure.
— Should I (or my company) switch to Claude? That depends on your situation. If coding and developer workflows are core, there's plenty of reason to seriously test Claude. But "it's number one so just do it" is a risky approach — judge by actual fit for your work, not the brand ranking. It's too early to declare either side the answer, and where possible I'd recommend trying several models in parallel before picking.
Sources
- Anthropic finally beat OpenAI in business AI adoption — VentureBeat
- Anthropic overtakes OpenAI in B2B adoption for the first time — The Decoder
- Ramp AI Index — Anthropic beats OpenAI on business adoption
- Anthropic overtakes OpenAI in workplace AI adoption — Axios
- Anthropic vs OpenAI business adoption analysis — FourWeekMBA
Numbers and criteria are as of announcement and may change. Investment calls are yours to make!
출처
관련 기사

Anthropic Just Toppled OpenAI in Revenue – Here's What That Actually Means

HumanX 2026: 'Claude Code Has Become a Religion' — The AI Industry Just Shifted

Anthropic Captures 70%+ of New Enterprise AI Deals vs OpenAI, Per Ramp Data
AI 트렌드를 앞서가세요
매일 아침, 엄선된 AI 뉴스를 받아보세요. 스팸 없음. 언제든 구독 취소.