OpenAI has begun testing advertisements in ChatGPT for users on its free and lowest-cost Go tier in the United States, marking a significant shift in how the company plans to monetise its flagship product. According to OpenAI, the trials place clearly labelled sponsored content beneath chatbot responses and are limited initially to a subset of U.S. accounts.

The company says the ads will be contextually relevant to the conversation, so a user discussing meals might see an offer from a grocery delivery service, and that ad placement will not alter the model's answers or permit direct access to personal conversation data for advertisers. OpenAI has also said it will exclude ads from sensitive topics and from users under 18.

The move has prompted immediate pushback from rivals and raised questions about the role of advertising inside AI assistants. Anthropic has positioned itself in direct contrast by pledging to keep Claude free of advertising, arguing that sponsored messages would undermine the assistant’s role as a space for focused work and reflection. The company reinforced that stance with marketing that mocked the prospect of conversational assistants pivoting into product pitches.

Anthropic has also expanded features available to free users, such as file creation, connectors and "skills", as part of a strategy to attract those who prefer an ad-free experience, signalling a commercial bet that users will pay or tolerate reduced functionality rather than accept ads. Industry coverage notes the timing of those moves directly followed OpenAI’s announcement.

OpenAI officials emphasise the tests are a measured attempt to balance wider access with commercial sustainability, including the introduction of ChatGPT Go at a modest monthly price point. Observers say the company faces mounting costs to deliver increasingly capable models and that advertising is one route to fund large-scale development while keeping entry-level access broadly available.

Beyond commercial calculations, the dispute has reopened wider debates about design choices for AI systems: whether assistants should remain neutral tools or be monetised through embedded advertising, and how safeguards can prevent the reappearance of engagement-driven harms associated with social media. The contrasting approaches taken by OpenAI and Anthropic offer a clear test case for how the market and regulators might shape the future of conversational AI.

Source Reference Map

Inspired by headline at: [1]

Sources by paragraph:

Source: Noah Wire Services