ChatGPT to start showing users ads based on their conversations

OpenAI is preparing to introduce advertising into ChatGPT, marking a significant shift in how the AI service is monetised. The company said it will begin testing ads within the free version of ChatGPT for logged-in adult users in the United States, positioning advertising as a new revenue stream alongside subscriptions.

The move comes as OpenAI continues to scale its infrastructure and services to support a rapidly growing user base. ChatGPT now attracts around 800 million monthly users, placing increasing financial pressure on the company to fund long-term investment in computing capacity and AI development. Advertising is being explored as a way to diversify income beyond subscriptions and enterprise contracts.

As part of the rollout, OpenAI is also introducing a new subscription tier called Go, priced at US$8 per month. The Go plan offers enhanced features such as longer memory and additional image generation capabilities. Unlike higher-tier subscriptions, Go users will still see ads, while Plus, Pro and business customers will remain ad-free.

A new monetisation strategy for a fast-growing platform

OpenAI has previously relied on paid subscriptions, developer tools and enterprise offerings to generate revenue. The introduction of ads reflects the scale of investment required to sustain large language models and associated infrastructure. The company has committed to spending an estimated US$1.4 trillion on AI infrastructure over the next eight years, underlining the urgency of expanding revenue sources.

In late 2024, OpenAI projected that it would reach approximately US$20 billion in annual revenue by the end of 2025. While subscription growth has contributed to this figure, advertising offers access to a broader segment of users who rely on the free version of ChatGPT and may not be willing to upgrade to paid plans.

The company has already taken steps towards commercial transactions within ChatGPT. Last year, it launched Instant Checkout, a feature that allows users to purchase products directly from retailers such as Walmart and Etsy without leaving the chatbot. It has also expanded into areas such as health and education tools, signalling an ambition to embed ChatGPT more deeply into everyday digital activities.

How ads will work and what users can expect

During the testing phase, ads will appear at the bottom of ChatGPT responses and will be clearly labelled as sponsored. OpenAI has emphasised that advertising will not influence how the model answers questions, stating that responses will continue to be driven by what is considered objectively useful rather than commercial considerations.

Advertising is expected to be contextually relevant, using the content of user interactions to surface products or services that align with recent queries. For example, a user seeking help with travel planning could see ads related to accommodation or local experiences. This approach mirrors broader trends in personalised digital advertising, where relevance is prioritised over broad targeting.

At the same time, OpenAI has outlined limits on how advertising data will be used. The company said it will not sell user conversations or personal data to advertisers, and users will be able to disable ad personalisation based on their chats. Ads will also be excluded from conversations involving regulated topics such as health, mental health and politics.

Privacy, trust and wider industry implications

Introducing ads into conversational AI raises questions around user trust and data sensitivity. Chatbot interactions can involve personal or intimate topics, increasing scrutiny over how data is interpreted and monetised. OpenAI has acknowledged the importance of maintaining confidence in the platform, particularly as ChatGPT becomes more integrated into daily decision-making.

The company has also said that ads will not be shown to users under 18. To enforce this, OpenAI relies on AI-based age estimation using conversation patterns and usage behaviour. This approach reflects broader challenges faced by platforms seeking to balance personalisation, safety and regulatory compliance.

OpenAI’s decision follows similar moves across the AI sector. In December, Meta began using information from user interactions with its own AI chatbot to improve ad targeting. As generative AI tools become more mainstream, advertising is increasingly viewed as a natural extension of existing digital business models.

While advertising offers significant revenue potential, it also increases pressure on AI providers to ensure responsible recommendations. OpenAI has faced legal challenges in the past, including lawsuits alleging that ChatGPT promoted harmful behaviour. The addition of commercial content heightens the need for safeguards to prevent the promotion of unsafe or inappropriate products.