OpenAI collaborates with Smartly to introduce conversational advertisements within ChatGPT.

      It only took six weeks. On February 9, OpenAI activated advertisements for free-tier users of ChatGPT in the United States. By the end of March, the company revealed that the initiative had surpassed $100 million in annualized revenue, attracted over 600 advertisers, and reached less than one-fifth of eligible users. OpenAI is now seeking external assistance to enhance the ads significantly. According to Business Insider, the company is collaborating with Smartly, an advertising automation platform based in Helsinki, to develop conversational, interactive ad formats for ChatGPT that engage with users rather than merely appearing next to responses.

      This partnership signifies a quiet yet notable shift from the static, labeled placements that OpenAI initially implemented when it announced its advertising plans in January. In these early tests, a sponsored card could show up below a product-related query, such as an air fryer ad after a question about kitchen appliances or a hotel advertisement following a travel query. The new conversational formats, as described by Smartly in their collaboration overview, will enable users to click on an ad and engage in a chatbot experience that offers customized recommendations, effectively creating a secondary dialogue within ChatGPT’s interface.

      Smartly, which reported approximately $101 million in revenue in 2025 and is valued at around $300 million, is primarily known for helping brands optimize their campaigns in real-time across platforms like Meta, Google, TikTok, and Snapchat. Its role in the ChatGPT collaboration seems to focus on performance optimization, adjusting creative content and targeting dynamically based on user interactions with the new ad formats. Additionally, OpenAI has partnered with Criteo, a French commerce-media company, which has connected about 17,000 advertisers to ChatGPT’s inventory since March 2. Conversations have reportedly taken place with The Trade Desk regarding further scaling ad sales, although no agreements have been made public.

      The commercial infrastructure is being rapidly established. In late March, OpenAI recruited David Dugan, a former vice president at Meta who led its global clients and agencies division for over a decade, to oversee its new global advertising solutions team. Dugan reports to COO Brad Lightcap and joins a leadership team that features Fidji Simo, the former Facebook executive currently serving as CEO of OpenAI’s applications division, and Denise Dresser, the former Slack CEO appointed as chief revenue officer in December 2025. These hires indicate a shift from experimentation with advertising to the development of a permanent sales organization around it.

      OpenAI has informed investors that it anticipates ChatGPT consumer revenue will surpass $17 billion in 2026, with advertising making up a significant portion of income from users who do not pay. The numbers suggest a promising outlook: the company claims over 800 million weekly active users, but only about 5 percent subscribe to paid plans. Self-serve advertising tools, eliminating the current $200,000 minimum commitment and allowing access for small and medium-sized businesses, are set to launch this month. Pilot programs in Canada, Australia, and New Zealand are expected to follow, with broader international expansion planned throughout the year.

      The initial pricing has been aggressive. ChatGPT ads were introduced at around $60 per thousand impressions, a premium that OpenAI justifies by citing early performance statistics: Criteo reports that users directed from large language models convert at 1.5 times the rate of other channels. Nearly 80 percent of small and medium-sized businesses have reportedly expressed interest in advertising on the platform, a statistic that, if it translates into actual spending after the launch of self-serve access, could represent a remarkably swift increase.

      The privacy framework surrounding these ads will be crucial in determining whether the conversational formats succeed or spark a backlash that undermines ChatGPT's core appeal. OpenAI asserts that conversations remain confidential and are never shared with advertisers, who only receive aggregated performance metrics like views and clicks. Ads are prohibited from appearing alongside health, mental health, and political topics, and users under 18 do not see them. However, these safeguards have not quelled all criticism. Zoe Hitzig, a former OpenAI researcher, has publicly claimed that the company is creating an economic engine whose incentives might eventually override its own rules, highlighting that ChatGPT contains an unprecedented archive of deeply personal conversations. Purchase-tracking data can be relayed back to OpenAI when users click on an ad and make a purchase, and the updated privacy policy allows for contact-syncing features that could process a user's phone number even without their consent.

      Competitors have quickly seized this opportunity. Anthropic, the creator of the Claude chatbot, aired Super Bowl ads that implicitly contrasted its ad-free model against OpenAI's approach. The messaging was clear: one company sees its users as the product, while the other does not. Whether this distinction matters to the hundreds of millions of people who use ChatGPT for free remains uncertain, but the reputational risks are notable for a company that has presented itself as the responsible steward of artificial general intelligence.

      The conversational