Writing

• AI in Marketing
• 23 September 2026

AI in marketing statistics 2026: UK and global data, sourced

The headline AI in marketing statistics for 2026 show near-universal adoption among marketers (HubSpot puts it at 86.4% of teams, Salesforce at 75%) and real money moving, with Gartner finding CMOs put 15.3% of marketing budgets into AI. Proven returns are much thinner: McKinsey finds only 6% of organisations are AI high performers.

This page was last updated September 2026. Every figure below comes from a report or survey page I opened, with its sample and date, and UK numbers are kept separate from global ones. The single most sobering number is from a 2026 Ipsos study led by Adam Peruta of Syracuse University: AI-made versions of 20 real video ads scored 14% lower on short-term sales potential and 17% lower on long-term brand equity than the human-made originals, even though 3,000 US viewers struggled to tell them apart.

How widely is AI used in UK businesses and UK marketing?

What the UK AI marketing adoption figures mean

The 35% versus 16% gap looks like a contradiction until you read the footnotes: ONS counts firms with ten or more staff in June 2026, while DSIT counts every business including micro firms, with fieldwork more than a year earlier. Most adoption is also shallow, and 1.6 technologies per adopter is somebody using ChatGPT and somebody else using Canva’s image generator, not an AI operating model.

The number marketers should care about is DSIT’s 72%. Among UK firms using or planning to use AI, marketing is joint first with admin as the place it goes, which makes sense because a first draft of a LinkedIn post is lower risk than letting a model near the payroll. The ONS gap between 55% of employees and 35% of businesses points the same way: people are using AI whether or not their employer has a policy, so plenty of UK marketing teams already have an AI strategy, it just lives in individual browser tabs.

What percentage of marketers use AI globally?

• 86.4% of marketing teams say they use AI in at least a few marketing areas. HubSpot 2026 State of Marketing, 1,500+ marketers across regions and company sizes (HubSpot, 2026).
• 75% of marketers have adopted AI, yet 84% admit running generic campaigns. Salesforce State of Marketing, tenth edition, 4,450 marketing decision-makers in North America, Latin America, Asia-Pacific and Europe, surveyed 8 October to 17 November 2025 (Salesforce, February 2026).
• 97% of marketing leaders use AI in their daily creative work. Canva with The Harris Poll, 1,415 marketing leaders at organisations with 500+ employees across the US, UK, Australia, France, Germany, Japan and India (Canva, 2026).
• Nearly nine in ten respondents report regular use of AI in at least one business function, and 56% use it in three or more functions, up from 51%. McKinsey State of AI 2026, 1,719 participants in 97 countries, 4 May to 8 June 2026 (McKinsey, August 2026).
• 42% of CMOs use generative AI only as an assistant for individual tasks, and only 8% run campaigns in which multiple AI agents operate autonomously. BCG, 300 CMOs across B2C and B2B plus 50 structured interviews (BCG, June 2026).

Surveys of marketers about AI tend to find that marketers use AI, in the same way surveys of runners find that people own trainers. The definition of “use” stretches from “we’ve automated our paid social bidding” to “someone asked ChatGPT for a subject line once in March”. BCG’s split is the honest one because it separates the two: 42% of CMOs are still in assistant mode, and fewer than one in ten run campaigns where several agents work on their own. I’ve set out the difference in generative vs agentic AI in marketing.

How much are marketers spending on AI?

Put Gartner’s numbers next to each other and the arithmetic is slightly alarming. Total marketing budgets are flat at 7.8% of revenue, 56% of CMOs already say they can’t afford their strategy, and yet 15.3% of that flat pot is going to AI. That money is coming out of something, and my guess is the lines that are hard to defend in a budget meeting, like research: the IPA Bellwether puts UK market research budgets at a net balance of minus 4.1% in Q2 2026. If so, we are funding the new thing by cutting the tools that would tell us whether it works, which is a very marketing way to run an experiment.

Is AI in marketing delivering ROI?

• 37% of respondents attribute at least some EBIT impact to AI use, about the same share as last year. McKinsey State of AI 2026, 1,719 participants.
• 6% are AI high performers, defined as attributing 5% or more of EBIT to AI and reporting “significant” value. Same McKinsey survey.
• 31% of B2C CMOs and 20% of B2B CMOs say their agentic marketing transformation is already having significant, measurable revenue impact. BCG, 300 CMOs, June 2026.
• Only 30% of CMOs report mature or fully developed AI readiness, while 70% call becoming an AI leader a critical goal for 2026. Gartner, 401 CMOs, January to March 2026.
• 26.5% of marketers say AI has significantly increased productivity; about a third say it saves their team 10 to 14 hours a week and another third say over 15 hours. HubSpot 2026, 1,500+ marketers.
• 67.5% of marketers say they know how to measure AI impact, up from 48% in 2025. Same HubSpot survey.

The adoption versus ROI gap

Nearly nine in ten organisations use AI, about a third can point to any profit effect, and roughly one in seventeen can point to a meaningful one. McKinsey’s respondents most commonly report revenue gains from AI in marketing and sales, which is the good news, but most of the evidence marketers themselves offer is time saved. Hours saved is an input. Nobody has ever been promoted because the brand tracker went flat 14 hours a week faster.

I’d be wary of HubSpot’s 67.5% who “know how to measure AI impact”, since confidence in measurement tends to rise faster than measurement. If you want to test your own numbers, the method I use is in how to measure the ROI of AI in marketing, and it starts with a baseline you record before the tool goes live, which most of us skip because we were too excited.

Does AI-generated advertising actually work?

In the study described by Adam Peruta in HBR (September 2026), Peruta, an associate professor at Syracuse University’s Newhouse School, co-led research with Ipsos in which 3,000 US consumers watched 20 video ads from major brands: half human-made originals, half made entirely with AI from briefs derived from those originals.

• AI-made ads were on average 14% weaker on short-term sales potential.
• They were 17% weaker on long-term brand equity.
• The AI versions of a Febreze product demonstration and a Herbal Essences beauty ad performed at least as well as the human originals on short-term sales potential.
• More than half of participants said a brand using AI in ads wouldn’t change how they felt about it, but a large majority said AI use should be disclosed.

AI did fine where the brief asked for a familiar structure (problem, product, reason to believe) and fell down where the idea depended on a joke, an emotional turn or a surprising point of view. Peruta’s Fiat example is the one that stuck with me: the brief said “energetic, defiant, playful, and rebellious”, and the AI took that literally and produced moody neon driving shots, while the human version turned the same “testing” idea into a joke. Which is roughly what happens when you give a junior a list of adjectives and no insight, so the lesson is partly about briefs.

The practical risk is that AI creative sails through review because it looks finished and nobody in the approval chain measures persuasion. We sign it off, it runs, the brand tracker softens a year later and we blame the media mix, the category, the pricing and probably the weather. If you’re using AI for creative, pre-test AI and human versions against the same brief before you scale, which is what I’d build into any AI marketing workflow.

What do consumers think about AI in marketing?

• 82% of ad executives believe Gen Z and Millennial consumers feel positive about AI-generated ads; 45% of those consumers actually do. The gap widened from 32 points in 2024 to 37 points. IAB (US) with Sonata Insights, 505 US consumers aged 16 to 43 and 104 US ad executives (IAB, January 2026).
• 20% of those consumers associate brands using AI ads with “manipulative”, against 10% of executives; 22% of consumers pick “forward-thinking”, against 46% of executives. Same IAB study.
• 78% of consumers say they’d rather see ads made by people, and seven in ten say AI-generated ads feel like they’re missing something; 68% say they don’t mind AI in advertising if it makes ads more helpful. Canva with The Harris Poll, 3,547 consumers across seven countries including the UK (Canva, 2026).

Ad executives are more than twice as likely as young consumers to see AI ads as forward-thinking, which is a bit like the chef being the only one in the restaurant who loves the new menu. For UK teams, the regulatory position is simpler than the sentiment: the ASA and CAP said in May 2025 that there is no blanket legal requirement in the UK to disclose AI use in ads, but the usual rules on misleading advertising apply whether AI made the ad or not. Consumer expectations are ahead of the rules, and the Peruta and Canva data both suggest disclosure costs less than people fear.

What are the main AI in marketing trends for 2026?

• 58% of IAB UK members are experimenting with or piloting agentic AI, a further 16% are scaling agentic systems, and 4% consider themselves fully “agent-first”. IAB UK, State of AI in Advertising, June 2026.
• 47% of advertisers don’t trust AI agents in advertising because decision-making isn’t transparent, rising to 67% among IAB UK members. Same report, as covered by Marketing Week.
• 74% of advertisers believe AI summaries are reducing traffic to brand websites. IAB UK, June 2026.
• Nine in ten clients describe AI skills as critical or very important when choosing creative and media agencies. IPA with Tracksuit, 200 senior brand-side decision-makers plus 20 interviews (IPA, July 2026).

The agent numbers need a pinch of salt, because “experimenting with agentic AI” can mean a single Zapier flow somebody built on a Friday. The trust gap is more telling, because IAB UK’s members are the ad tech and media businesses closest to these systems, and 67% of them don’t trust the agents. I’d act on the AI search number first: 74% believing traffic is falling is only a belief, but it is about your own channels and cheap to check in your analytics. For where AI sits in planning and buying, see AI in marketing and advertising.

How to read an AI in marketing report before you quote it

A lot of AI in marketing research is published by companies that sell AI marketing software, which doesn’t make it wrong, but the questions are written by someone with a quarterly target. Before a stat goes into your deck, run it through this.

1. Find the primary source. If a number only appears on a “100 AI statistics” blog, it doesn’t go in.
2. Check who was asked. ONS’s headline 35% covers firms with ten or more staff; DSIT’s 3,500 includes micro businesses. Those produce very different adoption numbers.
3. Check the fieldwork date, not the publication date. Salesforce’s 2026 report was fielded in late 2025, and DSIT’s January 2026 report was fielded in spring 2025.
4. Separate self-reported benefit from measured outcome. “Saves 10 to 14 hours a week” is a perception; “14% lower sales potential” is a test result.

A prompt I use to pressure-test a report before citing it (paste the methodology section in with it):

Here is the methodology and headline finding from a marketing survey.
1. Who was surveyed, how many, where, and when was fieldwork?
2. What exactly was the question behind the headline figure?
3. Is the finding self-reported perception or a measured outcome?
4. What population can this figure fairly be generalised to?
5. Does the publisher sell a product the finding supports?
Answer only from the text provided. If something isn't stated, say "not stated".

The last line matters, because a model left alone will happily invent a sample size to be helpful. For the basics of what these tools are and aren’t, the what is AI in marketing guide is the starting point, and the ACTAsia case study shows what adoption looks like in a small charity team where nobody has a $15 million budget. If you only check one number from this page against your own data this month, make it the AI search one: pull organic sessions to your top ten landing pages for the last 18 months and see whether the 74% belief holds for you.

FAQs

What percentage of UK businesses use AI in marketing?

The best official figure comes from DSIT’s AI Adoption Research (3,500 businesses, fieldwork February to May 2025) found 72% of businesses using or planning to use AI apply it in marketing, tied with administration as the most common use. ONS separately found around 35% of UK firms with 10+ employees used AI in June 2026.

How much of the marketing budget goes on AI in 2026?

Gartner’s 2026 CMO Spend Survey of 401 marketing leaders in North America, the UK and Europe found CMOs allocate an average of 15.3% of marketing budgets to AI initiatives. Organisations with mature AI readiness allocate 21.3%. Total marketing budgets were flat at 7.8% of company revenue, which suggests AI spend is mostly being found inside existing budgets.

Do AI-generated ads perform as well as human-made ads?

Not on average, according to a 2026 Ipsos study co-led by Adam Peruta of Syracuse University and published in HBR. Across 20 video ads shown to 3,000 US consumers, the fully AI-made versions were 14% weaker on short-term sales potential and 17% weaker on long-term brand equity, although viewers struggled to tell them apart. Some simple, product-led ads matched the originals on short-term sales potential.

Where can I find reliable AI in marketing research and articles?

Start with primary sources that publish their methodology: ONS and DSIT for UK adoption, IAB UK and the IPA for UK advertising, McKinsey’s State of AI and the Gartner CMO Spend Survey for global business data, and HBR for research-led studies such as the Peruta and Ipsos ad test. Treat vendor reports from HubSpot, Salesforce or Canva as useful but check sample, fieldwork dates and question wording.

Sources

1. Office for National Statistics, “Artificial intelligence in UK businesses: 2023 to 2026”, 20 July 2026
2. Department for Science, Innovation and Technology, “AI Adoption Research”, 28 January 2026 (updated 13 February 2026)
3. IAB UK, “Powering Growth: The UK’s Hidden AI Engine”, 14 May 2026
4. PPC Land, “IAB UK: 95% of UK digital ad businesses already use AI”, May 2026
5. IAB UK, “One Third of UK Digital Ad Spend Will Be AI-Driven by 2030, IAB UK Forecasts”, 2 June 2026
6. Marketing Week, “AI expected to drive £18bn of UK digital ad spend by 2030”, June 2026
7. IPA, “UK companies revise marketing budgets up in Q2 2026 despite tough economic conditions” (Bellwether Report Q2 2026), 16 July 2026
8. IPA, “New IPA/Tracksuit report reveals what clients want from their agencies”, 8 July 2026
9. Gartner, “Gartner 2026 CMO Spend Survey Finds CMOs Allocate 15.3% of Marketing Budgets to AI, But Only 30% Are Ready to Scale AI Capabilities”, 11 May 2026
10. McKinsey, “The state of AI in 2026: On the road to ROI”, 25 August 2026
11. BCG, “Moving the Agentic Marketing Transformation from Illusion to Reality”, 15 June 2026
12. BCG, “Mind the Marketing Gap: Most CMOs Say AI Is Transforming Marketing, But Few Are Using It to Transform Their Own Function”, 15 June 2026
13. Salesforce, “75% of Marketers Have Adopted AI Yet Still Use It To Send One-Way, Generic Campaigns” (State of Marketing, tenth edition), 19 February 2026
14. HubSpot, “2026 state of marketing: Data from 1,500+ global marketers”, updated April 2026
15. Canva, “AI is in. Now comes the hard part: earning consumer trust” (2026 State of Marketing and AI Report, with The Harris Poll), 2026
16. IAB (US) with Sonata Insights, “The AI Ad Gap Widens”, 15 January 2026
17. ASA and CAP, “Disclosure of AI in advertising: striking the balance between creativity and responsibility”, 29 May 2025
18. Harvard Business Review, Adam Peruta, “Research: AI-Generated Ads Perform Worse Than Human-Made Ones, Even When Customers Can’t Tell Them Apart”, 3 September 2026