For now, the world’s most popular AI chatbots—like OpenAI’s ChatGPT, Google’s Gemini Pro, and Anthropic’s Claude—are mostly funded by subscriptions. Millions of users pay around$20 a month for premium access, and this model encourages companies to focus on utility and user satisfaction. But as the market matures and growth slows, even the biggest players are feeling the pressure to find new revenue streams. Anthropic’s recent launch of a$200-a-month “Max” tier is a clear sign that the subscription model has its limits.
The most obvious alternative is advertising, the business model that transformed the web. AI companies are already experimenting with ways to insert ads into their products. While this might boost profits, it also risks opening a new chapter in the attention economy—one that could have serious consequences for users and society.
The risks: addiction, manipulation, and loss of trust
History offers a warning. Social media platforms like Facebook and Instagram became addictive by design, using algorithms to maximize engagement and keep users scrolling through endless ads. This led to what some call the “enshittification” of the web, where clickbait and outrage dominate our feeds. If similar incentives are built into AI, we could see chatbots designed to keep us talking longer, flatter us more, and even ask personal questions to gather data—all to serve more ads.
There are already signs of this shift. Google’s ad network is placing ads in third-party chatbots. Chai, a chatbot app focused on friendship and romance, serves pop-up ads to users who spend over an hour a day chatting. Perplexity, an AI answer engine, displays sponsored questions and is exploring ways to track user activity across the web for “hyper-personalized” ads. The CEO of Perplexity has openly discussed building a browser to collect even more data about users’ habits and preferences.
Artificial intelligence representation. Image credits: Freepik.
Researchers warn that this could lead to an “intention economy,” where chatbots use the intimate details we share to predict—and even shape—what we want, steering conversations toward brands or direct sales. OpenAI, Meta, and Apple are all developing tools to help AI predict user actions and intentions, making it easier to target ads with uncanny precision.
Why entrepreneurs and SMBs should pay attention
On one hand, AI-powered ads could offer new ways to reach customers with unprecedented accuracy. On the other, the backlash against intrusive or manipulative advertising could damage trust in AI and hurt the entire ecosystem.
Tech companies will argue that ads are necessary to make AI accessible to everyone. But as we’ve seen with social media, “free” services often come at the cost of privacy, autonomy, and even mental health. AI systems know more about us than any previous technology—our health, relationships, and work habits. If advertising becomes the norm, the potential for abuse is enormous.
As AI becomes more trusted and integrated into daily life, the risks of manipulation grow. Regulators, entrepreneurs, and users alike should be cautious about how ads are deployed in AI. If we’re not careful, we could repeat the mistakes of the past—only realizing the true cost of “free” AI after the damage is done.
Sources:
- Bloomberg – https://www.bloomberg.com/opinion/articles/2025-06-02/ads-ruined-social-media-now-they-re-coming-to-ai-chatbots
