ChatGPT Ads Prompt Senator Markey to Take Action

Imagine a digital assistant to whom you ask your most private questions, perhaps to unwind from the day's stress, suddenly trying to sell you a product. This exact scenario is about to become reality with the ad tests OpenAI announced for ChatGPT. And this move has promptly drawn the first political reaction from across the Atlantic.

Democratic Senator Ed Markey from Massachusetts is very clear in his official letter to OpenAI CEO Sam Altman. He states that 'sponsored' content to be placed beneath free users' chats raises serious concerns regarding consumer protection, privacy, and especially the safety of young users. Markey's concern is not limited to OpenAI. Similar letters have also been sent to AI chatbot developers Anthropic, Google, Meta, Microsoft, Snap, and even Elon Musk's xAI. They have all been given a period of time to respond.

Is Emotional Bond an Opening for Ads?

So what is Markey's real fear? It essentially lies in the nature of the relationship we build with these technologies. You search for a product in a classic search engine, an ad appears, and you easily understand it's a marketing message. However, the dialogue you establish with an assistant like ChatGPT or Google Gemini can be much more personal, even intimate. People sometimes share their daily problems, health concerns, or relationship issues on these platforms.

It is precisely at this point that Markey issues a rather striking warning: Companies could target users by leveraging this 'emotional bond' and 'trust relationship' fostered by their own systems. In other words, an AI you confide in has the potential to place ads without you noticing. OpenAI's statement on the matter does not fully alleviate this concern either. The company says 'chat interfaces create the opportunity to go beyond static messages and links.' What does this mean? We might encounter marketing messages that are much more sophisticated than a traditional banner ad, perhaps hidden within the natural flow of the conversation.

OpenAI has announced it is setting some limits for now. It will not show ads to users under 18. Ads will also not appear in chats on sensitive topics like physical health, mental health, or politics. But there is a point Markey questions: Will the data collected from these sensitive conversations not be used to personalize the ad you see in your next chat? The answer to this question is not clear.

Are Personal Thoughts Now a Meta?

The privacy issue is another dark corner of this debate. In his letter, Markey emphasizes that people's 'personal thoughts, health questions, family matters, and other sensitive information' should never be used for targeted advertising. This is essentially a new reflection of general internet privacy debates in the age of AI. The difference here is that the data is much deeper and contextual.

The statement from OpenAI spokesperson Liz Bourgeois reflects the company's perspective: 'Any advertising we explore is about supporting the resources needed to keep this technology freely accessible for ordinary people.' So, the model is this: Offer a free service, let the user base grow, then sustain this free model with ad revenue. This resembles the path social media platforms have followed for a decade. Bourgeois says the company will be 'guided by transparent principles designed to preserve trust' and will not be optimized solely to 'maximize engagement.'

However, history shows us that when ad revenue is involved, maintaining the balance between 'engagement' and 'trust' is not easy at all. We have seen news sources manipulate their algorithms, social media feeds steer us towards certain content. What about chatbots? Their capacity to steer us could be much stronger. Because you have an entity in front of you that claims to understand you and answers your questions.

This situation points to an area where regulations need to swiftly come into play. Markey's move is likely the first step towards broader AI and digital advertising regulation. The European Union's AI Act already sets some limits, but the legal gap on this issue is more pronounced in the US. Ultimately, AI has ceased to be just a technology issue; it has transformed into a complex area where consumer rights, ethics, and policy intersect. The arrival of ads in ChatGPT might seem like a small step, but it could be a major turning point that determines the future of this digital relationship.