AI is moving attention and ad money toward the largest platforms.

Digital platforms and publishers make money from attention. They draw an audience, then earn from ads, subscriptions or fees. Most smaller ones get much of that audience from search engines and large social apps. AI is changing both.

Search sends fewer visitors. When Google shows an AI summary at the top, people click a link to a website in 8% of visits. Without a summary, they click in 15%. Only about 1% click a link inside the summary. The study followed about 900 American adults for a month. Google disputes the study's method. But for a site that lives on search traffic, even part of that drop matters.

AI answers are a new front door. More people now ask a chatbot instead of searching. When a chatbot answers, it may use your content without sending anyone to your site. Being named and linked in those answers matters more each year.

Your content has a new buyer. AI companies need large amounts of good writing, video and data. Some pay publishers and platforms to use it. That can be a new source of money, or a way of paying for the traffic you lose. The terms decide which.

Ad money is flowing to the largest platforms. Meta's ad revenue grew 27% last quarter. The average price of an ad rose 12%. Meta says its AI tools find buyers better, and more than 9 million small businesses use its AI to make ads. When the largest platform gets better at selling ads, advertisers move money there. Smaller platforms must prove their ads work as well.

AI is expensive, even for the winners. Meta's operating margin fell from 43% to 31% in a year. Part of the reason was the cost of data centers and AI. A smaller platform that adds AI features will pay for them too, often to the same few AI companies.

Content is cheaper to make, and harder to trust. AI can write posts, make images and dub videos for very little. That helps a small team publish more. It also floods platforms with cheap content, and someone has to sort the real from the fake.

New rules apply in Europe. From August 2, 2026, a chatbot must tell people it is AI. Deepfakes must be labeled. Tools already on the market have until December 2 to mark AI content in a way machines can read. Fines can reach €15 million or 3% of worldwide revenue.

Time saved is not money saved. AI that writes content faster pays only if the extra content earns more than it costs, or the team spends less.

Here is what that looks like in practice. A publisher earns $80 million a year, and 40% of its visitors come from search. If AI summaries cut those visits by a third, it loses about 13% of its traffic. At the same ad rates, that is roughly $10 million a year. Nothing about its content changed.

What should a platform's leaders do first? Three plain things.

1. Measure how much traffic comes from search, and track it every month.
2. Build ways to reach your audience directly: email, apps, subscriptions and community.
3. Check your products against the new European rules, especially chatbots and AI images.

None of this needs a new team or a large budget. It needs one leader who owns each answer.

The full guide shows where AI touches the money, what has actually happened, who gains and who loses, and what to ask before you spend.