ChatGPT users may soon start seeing ads in their chats, as OpenAI announced on Monday that it’s officially beginning to test ads on its AI platform. They’ll appear as labeled “sponsored” links at the bottom of ChatGPT answers, but OpenAI says the ads “do not influence the answers ChatGPT gives you.”

Currently, ads will only show up for users on the free version of ChatGPT or the lowest-cost $8 per month Go plan. Users in the Plus, Pro, Business, Enterprise, and Education plans won’t see any ads, so anyone who wants to avoid them has to pay at least $20 per month for the Plus subscription. There is one loophole — OpenAI notes that users can “opt out of ads in the Free tier in exchange for fewer daily free messages.”

Users on the Go tier can’t opt out of seeing ads, but users on both the Free and Go plans can dismiss ads, share feedback on ads, turn off ad personalization, turn off the option for ads to be based on past chats, and delete their ad data. According to OpenAI, advertisers will only get data on “aggregated ad views and clicks,” not personalized data or content from users’ ChatGPT conversations.

Additionally, not all users and chats will be eligible for ads, including users under 18 and conversations on certain sensitive topics “like health, mental health or politics.” Even adult users on the chatbot’s Free and Go plans might not immediately start seeing ads, since the feature is still in testing.

The introduction of ads on ChatGPT was initially announced a few weeks ago, and its launch was rumored just a few hours before OpenAI’s official announcement on Monday. Ahead of the launch, rival Anthropic took a jab at ads on ChatGPT in its Super Bowl commercial on Sunday. The ad drew backlash from OpenAI CEO Sam Altman, who claimed it was “clearly dishonest,” and Anthropic ended up changing the ad’s tagline to make it less directly about OpenAI, stating instead: “There is a time and place for ads. Your conversations with AI should not be one of them.”