Advertising and marketing agencies increased their spending on AI tools threefold between January and June 2026. The finding is drawn from anonymized card transaction data covering Wallester Business clients classified under Advertising and Marketing. Spend on named AI tools rose from an estimated 1.02% of total card turnover in January to 3.01% in June.
The figure reflects actual card spend rather than self-reported usage, which makes it a firmer measure of adoption than much of the current commentary on AI in advertising.
A Gradual Increase, Not a Single Spike
The increase was gradual rather than sudden. AI spend as a share of total turnover rose every month: 1.02% in January, 1.34% in February, 1.73% in March, 2.28% in April, 2.94% in May, and 3.01% in June.
This consistent upward trend, rather than a single adoption event followed by a plateau, suggests an ongoing shift in how agencies allocate software budgets. Clearly, this is not a short-term spike tied to a single product launch or news cycle.
A Market Concentrated in Two Vendors
Breaking the same spend down by vendor reveals a market far more concentrated than the growth curve alone would suggest.
OpenAI accounts for an estimated 49% of all named AI spend in this segment over the first half of the year, and Anthropic for 43%, together 92.5%.That concentration lines up with how the broader market already treats the two companies. Anthropic has reportedly filed confidentially for an IPO targeted as early as October 2026, and OpenAI has reportedly been weighing a delay of its own listing into 2027, both further along toward public markets than any other AI company. Investors and ad agencies appear to be reaching the same conclusion from two different directions: these are the two clear leaders.
Perplexity ranks a distant third, at under 4%. Every other named vendor in the data, including Midjourney, Runway, ElevenLabs, Character.AI, Grammarly, Jasper, Stability AI, and Mistral, accounts for roughly 4% combined, spread across eight companies.
In other words, this is not a fragmented or experimental market but a market already defined by two vendors, with a long, thin tail behind them.
What the Spending Is Buying
Grouping the same spend by function rather than vendor clarifies the pattern further. General-purpose assistants, ChatGPT, Claude, and Perplexity, account for approximately 96% of the total. Image and video generation tools, along with voice tools such as ElevenLabs, account for approximately 3.5%. Dedicated writing assistants such as Grammarly and Jasper represent a negligible share.
This distribution indicates how agencies are currently allocating this budget. Most of it supports research, drafting, and reasoning, rather than the generation of a finished image, video, or voice track.
A substantial share of agency work takes place before the final creative asset is produced: briefs, pitch decks, client communication, media plans, and first drafts. This is the stage at which general-purpose assistants are typically used, and the spending data is consistent with that.
Context Behind the Numbers
Several things are worth stating plainly. The figures are based on anonymized, aggregated Wallester Business transaction data for clients classified under Advertising and Marketing.
The data captures only tools billed as identifiable merchants on card transactions, so spend routed through annual contracts, resellers, or procurement processes is not reflected here, meaning actual AI spend in this segment is likely higher than these figures indicate.
The figures measure total spend rather than the number of agencies using each tool, so this is a measure of intensity of use rather than an adoption rate. Finally, the data reflects Wallester’s own Advertising and Marketing client base rather than the industry as a whole. But it should be read as a verifiable signal from a real slice of the market.
Worth noting too: AI tools still sit well behind the segment’s largest cost, Advertising Services, at roughly two-thirds of total turnover. That’s not surprising. Advertising and marketing companies spending most of their budget on ads is simply how the business works.
In any case, within this context, two conclusions hold: spending on AI tools in this segment is increasing consistently month over month, and at present, that spending is concentrated in two vendors and in tools designed to support reasoning and drafting, rather than tools designed to produce finished creative output.
Where Wallester Business Fits In
Software spend tends to accumulate quietly within most agencies. Different employees subscribe to different tools using different cards, and the full picture typically only becomes visible once the monthly total appears on someone’s desk.
Wallester Business is built to bring that kind of scattered spending into one system, and getting started doesn’t require a subscription.
- Issue a separate virtual or physical card for every tool, team member, or client account, so spend is traceable from the first transaction
- Restrict a card to a single merchant, so a card issued for one AI subscription cannot be used elsewhere
- Apply spending limits and real-time controls per card, so a forgotten free trial does not quietly become a recurring charge
- Hold and pay in multiple currencies with no FX fees, relevant given that most AI vendors bill in US dollars
- Attach receipts and documents directly to each transaction, and sync everything with Xero or QuickBooks
- View every active card and its spend from one dashboard, on the web platform or the mobile app
- Open a business account and start issuing cards for free. Wallester Business is built around a Free to Start model, so there’s no cost to get into the system and begin managing spend
Wallester Business runs on Wallester’s own card infrastructure, not a reseller stack, and Wallester was named the fastest-growing fintech in Europe in this year’s Financial Times FT1000 ranking.
