How AI Is Changing Digital Marketing: Real Examples

TL;DR

• Adoption is real but uneven: as of 2025, 20.0% of EU enterprises with 10+ employees used AI, and large firms were more than three times as likely to as small ones.
• Ad buying is now largely automated: Google Performance Max and Meta Advantage+ decide bids, placements and creative combinations from the goals and assets you supply.
• Generated content now comes with rules: the EU AI Act’s Article 50 adds marking and disclosure duties for synthetic content.
• Discovery is shifting to AI answers: in a Pew study, people clicked a result link less often when Google showed an AI summary.
• Some things have not changed: consent law, the need for a real offer, and the fact that a model optimises for whatever you tell it to measure.

Contents

1. Who actually uses AI in marketing
2. Automated ad buying and creative
3. Generative content and labelling
4. AI search and discovery
5. Personalisation and its limits under GDPR
6. Measurement with less data
7. What has not changed and the risks
8. FAQ

Talk about “AI in marketing” used to mean predictions: smarter segments, chatbots, one day. That day arrived in a less cinematic form. AI now sits inside the ad platforms, the content workflow, the search results and the analytics reports. Below are the documented changes, described from official sources, plus the parts that stayed exactly as they were.

Who actually uses AI in marketing

Surveys of businesses give a sober picture. According to Eurostat, 20.0% of EU enterprises with 10 or more employees used AI technologies in 2025, up from 13.5% in 2024. Size matters a lot: Eurostat’s statistics explained page reports that in 2025 “17% of small enterprises, 30.36% of medium enterprises and 55.03% of large enterprises used AI”. Among enterprises that use AI, “34.70% of enterprises using AI technologies used these software or systems for marketing or sales”, ahead of the 31.05% that used it to organise business administration or management.

Large-company surveys show higher numbers, which is expected given who answers them. In McKinsey’s State of AI (2026 edition, survey of 1,719 participants run May–June 2026), “nearly nine in ten respondents report regular use of AI in at least one business function”, and “revenue gains, meanwhile, are most often attributed to the use of AI in marketing and sales.” Yet only 37% of respondents attributed at least some EBIT impact to AI use, about the same share as a year earlier.

The gap between “uses AI” and “earns money from AI” is the most useful number here.

Automated ad buying and creative

The biggest practical change is that the large ad platforms now run most campaign decisions themselves. The advertiser supplies a goal, a budget, assets and some signals. The system does the rest.

Google describes Performance Max as “a goal-based campaign type” that can “access all Google Ads inventory, including YouTube, Display, Search, Discover, Gmail, and Maps”. Bids are set with Smart Bidding, where “real-time auctions bid based on your conversion probability.” On the creative side, “Google AI generates headlines and descriptions from your landing page content, domain, and current ads”, and generative tools “combine your assets and creatives to help generate engaging images and videos.”

Meta calls its version Meta Advantage, “our portfolio of AI-powered automated business tools”. When Meta launched Advantage+ shopping campaigns, it said they automate “up to 150 creative combinations at once”. In 2023 it said that “using ASC, marketers can automate their entire setup of their campaign, creative and placement of their ad”, and that it was building “text variation, background generation and image outcropping.”

The trade-off: less manual work, less visibility. The marketer’s job moves from operating the controls to feeding the system good inputs, accurate conversion data and clear limits.

Generative content and labelling

Text, images, video and voice can now be produced in minutes. That changed the cost of making content, not the cost of making it good or true.

Regulation followed. The EU AI Act’s Article 50 requires providers of generative systems to ensure outputs are “marked in a machine-readable format and detectable as artificially generated or manipulated.” Deployers of deep fakes must disclose that content “has been artificially generated or manipulated”, and similar disclosure applies to AI-generated text published to inform the public on matters of public interest, unless it “has undergone a process of human review or editorial control” with someone holding editorial responsibility. Under Article 113, these duties apply from 2 August 2026. A 2026 amendment, Regulation (EU) 2026/1744, left that date in place but gives providers whose generative systems were already on the market before it a four-month transitional period for the machine-readable marking duty. We explain the details in EU AI Act Article 50: AI labelling explained.

For marketers, the practical point is simple. A realistic AI-generated person or voice in an ad is the kind of content these rules target. Disclosure is cheaper than a correction.

AI search and discovery

People increasingly get answers before they get links. Google said in May 2025 that AI Overviews is “driving over 10% increase in usage of Google for the types of queries that show AI Overviews” in markets like the US and India, and launched AI Mode, which it calls “our most powerful AI search.”

What that does to clicks has been measured from the user side. Pew Research Center tracked the browsing of 900 US adults in March 2025. Users clicked a traditional result link in 8% of visits when an AI summary appeared, and in 15% when it did not. About 18% of searches in the study produced a summary.

So visibility now has two layers: being ranked, and being the source an AI system chooses to cite or summarise. How those systems pick sources is covered in How AI search chooses sources.

Personalisation and its limits under GDPR

The old promise of “one customer, one profile, real time” is technically easier than ever. Legally, in the EU, it is bounded. Under GDPR Article 21, where personal data is processed for direct marketing, the person has “the right to object at any time”, including to “profiling to the extent that it is related to such direct marketing.” Once they object, “the personal data shall no longer be processed for such purposes.”

Article 21(3) leaves no room for exceptions here. Add the usual GDPR basics (a lawful basis, transparency, data minimisation) and the realistic version of AI personalisation looks like this: work mostly with data people knowingly gave you, explain what you do with it, and make opting out easy. A model trained on data you were not allowed to use is a liability, not an asset.

Measurement with less data

Consent banners and browser restrictions mean marketers see fewer complete user journeys. Platforms answer with modelling. Google explains that when a user declines ads or analytics cookies, Consent Mode “adjusts the relevant Google tags’ behavior to not read or write cookies for advertising or analytics purposes”, which leaves “a gap in their measurement”. Conversion modelling then uses consented journeys so that “our models will assess attribution paths for the unconsented journeys”, and modelled conversions appear in reports.

This is a real shift. Some numbers in a dashboard are now estimates produced by the same company that sells the ads. That is not necessarily wrong, but it is worth knowing which figures are observed and which are modelled, and checking them against your own sales data.

What has not changed and the risks

The main risks are concrete. Generated text can contain errors stated with full confidence. Automated campaigns can place ads in contexts you would not choose. Modelled reporting can make results look steadier than they are. And because every competitor has the same tools, generic AI content tends to converge on the same output. Generous for the platforms, less so for anyone trying to stand out.

FAQ

About the Author

Vyara Stefcheva

Vyara Stefcheva is the founder and editor of Egorithms and an independent marketing expert. She follows how technology, AI and algorithms change the way people find, choose and buy, and writes about what that means for businesses and for the people who use them.

She has spent her career building marketing and business strategies for more than 300 brands across three continents. She built and exited a marketing agency and now works independently. She has taught more than 20,000 people through lectures, courses and corporate training, and speaks at conferences about AI and the future of the internet.

On her YouTube channel she tests new AI and marketing tools as they launch and gives each an independent verdict: adopt, watch or ignore.

See author's posts