How advertising influences the results of AI (LLMs)

Like traditional search engines (such as Google), a growing number of LLMs include ad-supported responses or subtly embed ads within their answer text. Advertisers are therefore paying to manipulate the results. And currently, there's no guarantee that these ads will be clearly identified. As long as they remain, LLM AIs will obviously give higher probability of inclusion to this advertising content than from so-called "organic" content, regardless of the effort content creators put into it.

And the investment from advertisers is likely to be unprecedented: the impact of advertising on a generative AI (LLM) is more insidious and far more powerful than that of a traditional search engine (like Google). It has been shown that engaging in a conversation with an LLM has as much persuasive power as a conversation with a human. This makes them powerful sales and persuasion tools.

This is more and more recognized by legislators worldwide. There is growing pressure on governments (especially in Europe) to invest more in open and free AI, where capitalism has less influence. In Ontario, an initiative is being organized to regulate AI.

Similarly, organizations like Law Zero are developing technical means to ensure that AI models do not threaten the integrity of their users.

Until this is regulated, the amount invested in advertising will probably have more influence than the quality of the content and the website.