Could publishers claw back lost traffic by blocking AI summaries or charging per citation?

This explores whether opting out of AI summaries or getting paid per citation could make up for the clicks AI Overviews are taking from publishers. The short answer is that this collection doesn't evaluate either remedy. It has no studies of opt-out mechanisms, licensing deals or pay-per-citation schemes. What it does have is unusually solid evidence on the shape of the loss, and that evidence suggests what any fix would have to overcome.

Start with the size of the problem, because the numbers agree across very different methods. A randomized field experiment found AI Overviews cut clicks to publisher sites by 39.8 percent, with no measurable gain in how satisfied users were with their searches Do AI Overviews actually reduce clicks to publisher sites?. Ahrefs' click data shows the top-ranked result losing about 58 percent of its clickthrough when an Overview appears How much are AI Overviews actually reducing organic search clicks?. News executives surveyed by the Reuters Institute expect search referrals to fall 40 to 43 percent over three years Will AI Overviews reduce search referral traffic to publishers?. Even Wikipedia, the source an Overview is most likely to draw on, lost around 5 percent of English-language search traffic compared with German and French editions Does AI search summaries divert traffic away from Wikipedia?.

The more useful finding for your question is about why clicks disappear. An eye-tracking study shows the Overview now sits in the spot that searchers look at first. Attention to the first-ranked result dropped from 31 percent to 9 percent, while people trusted the AI summary as much as the ranked links Where do searchers look when AI Overviews appear?. The Wikipedia study describes the same mechanism: the answer satisfies the query before anyone reaches the source. This matters for both proposed remedies. Opting out removes your content from the summary, but it doesn't remove the summary. The box can be filled from someone else's material and still capture attention, so opting out may protect your content without bringing your clicks back. Pay-per-citation works with this mechanism rather than against it: it accepts that the click is gone and tries to charge for the citation instead. In effect, it turns publishers into suppliers to the answer layer rather than destinations.

The finding you might not have expected to matter is the 'no quality gain' result. If AI Overviews made searching clearly better, there would be a strong argument for paying publishers to feed them. The experiment found users weren't measurably better off, which weakens the case that the traffic shift is a fair trade for readers. That leaves the question as one of bargaining power rather than value created. Whether opt-out or payment schemes can change that balance is something this corpus can frame but not answer. You would need policy or economics sources that are not in the collection yet.