# Financing AI: OpenAI Launches Advertising Business in Europe
Starting Monday, ChatGPT will include advertisements in Europe. Although OpenAI CEO Altman long opposed the idea, the company is now banking on exactly this approach to fund its AI operations. Here's what the implications are.
From next Monday, the chatbot ChatGPT will contain advertising. In 31 European countries — including Germany, Austria, Switzerland, France, Spain, Italy, Poland and the Scandinavian countries — there will be advertisements appearing within chat conversations. This is not entirely new, however: a first trial run with advertising content at OpenAI already began in the United States six months ago.
"OpenAI is not completely alone in this: Microsoft already displays ads in Copilot, and Google is integrating advertising into its AI-powered search," Alicia von Schenk, Professor of Economics of AI and Human Behavior at Julius Maximilian University of Würzburg, told tagesschau.de. Other providers, meanwhile, have chosen a different path: "Anthropic has explicitly announced that it will keep Claude free of advertising. And Perplexity did test sponsored follow-up questions, but has since ended that advertising experiment out of concern for user trust."
In order not to forfeit that trust, OpenAI has emphasized that advertisers will not have access to chat histories: "The privacy of the conversation is fundamentally protected when it comes to displaying ads. We do not share any information from conversations with advertisers. This is extremely important to us — also because some users confide very personal things to ChatGPT."
## Advertising to Be Labeled
Nevertheless, displaying advertisements is not entirely unproblematic for AI providers like OpenAI, explains expert von Schenk: "Many people also use ChatGPT because they expect a direct, supposedly neutral answer and do not want to navigate through advertisements the way they do with a search engine. Overly intrusive advertising could damage this advantage and thus user trust."
OpenAI therefore intends to make advertising content visually distinguishable from chat content. Dave Dugan, Vice President of Ads at OpenAI, said: "ChatGPT's response is clearly separated from the ads and distinctly delineated." A grey line is to appear below the actual response, "and below that line there may be an advertisement."
## Indirect Effects on AI Responses Possible
Karoline Ströhlein from the University of Regensburg, however, emphasized to tagesschau.de that in an AI model that also delivers advertising content, the interests of the advertiser are also taken into account — after all, they too are paying customers. But this could become a problem for regular users: "A recent study on advertising in AI chatbots finds indications that various models in conflict situations sometimes act in favor of commercial interests — for example by recommending sponsored but more expensive products."
Indirect effects cannot be ruled out either, says expert von Schenk: "Advertisements could push certain brands or solutions to the forefront, thereby changing what users ask next." This altered question would then also lead to altered answers: "Advertising could thus influence the course of a conversation even when it does not directly intervene in the generation of responses." Karoline Ströhlein shares this concern: "When a chatbot tells me: 'For your situation, I would recommend product A, because …', it comes across as personal advice." But therein lies the danger that "users may evaluate the output not necessarily as a neutral collection of information, but partly as a decision-making aid."
There is yet another factor to consider: "Of course, there will also be something here that we already know from search engines, namely optimization," Professor Andreas Moring of the International School of Management told tagesschau.de. This already exists with conventional search engines today: through search engine optimization, or SEO, websites are designed so that search engines can easily find, understand, and display them prominently for relevant search queries. "Something like this will now also develop for AI," said expert Moring.
OpenAI CEO long opposed to it
The fact that OpenAI has decided to run advertising at all likely has clear financial motivations — after all, company co-founder and CEO Sam Altman was strictly opposed to it for a long time. "OpenAI and Anthropic want to go public. That is why they need to explain and demonstrate to investors what their future business model will look like," explains Andreas Moring.
The problem for OpenAI, according to Moring, is the following: "OpenAI is targeting private consumers." Yet according to current estimates, more than 80 percent of the over one billion monthly users still use the free version of ChatGPT. The decision to introduce advertising is therefore also important for the planned IPO: "It is a signal to investors ahead of the stock market listing that the company is truly serious about making money," explains Moring.
Advertising also intended to cover costs
Particularly since OpenAI has high costs: in order for the AI to be used, data centers must be built, power supply must be ensured, chips must be purchased, and specialist personnel must be paid. "In addition to subscriptions, corporate clients, and API fees, advertising is intended to help finance free usage in particular," explains expert von Schenk.
Those who do not wish to see these advertisements can, according to Dugan, also switch to ChatGPT Plus or Pro. The Plus subscription costs just under 23 euros per month in Europe, while the considerably more extensive Pro plan costs as much as 229 euros per month. Lower-priced entry-level subscriptions are already available from OpenAI. However, the entry-level plan ChatGPT Go, which costs just under seven euros per month, is not ad-free.
