Sen. Ed Markey (D-MA) is pressing OpenAI about its move to bring ads to ChatGPT, and is asking several other companies whether they have similar plans. In letters to the CEOs of OpenAI, Anthropic, Google, Meta, Microsoft, Snap, and xAI, Markey writes that embedding ads into AI chatbots “raises significant concerns for consumer protection, privacy, and the safety of young users.”
OpenAI will start testing ads for free ChatGPT users in the coming weeks, and they’ll appear in the form of “sponsored” products and services at the bottom of conversations with the chatbot. The company says it will surface ads relevant to your chat, though it won’t show them to users under 18, or during conversations related to physical health, mental health, or politics.
Even with these safeguards in place, Markey says the addition of ads in ChatGPT and other AI platforms “represents a significant, and potentially dangerous transformation” in the ad industry, as a user’s “emotional connection” to the chatbot could allow companies to “prey on the very relationships their systems have fostered.” Markey also points to OpenAI’s statement that “conversational interfaces create possibilities for people to go beyond static messages and links,” potentially making it harder for someone to recognize what is and isn’t an ad in the future.
Markey brings up potential privacy risks as well, saying that AI companies must not use an individual’s “personal thoughts, health questions, family issues, and other sensitive information” for targeted advertising. And while OpenAI says it won’t show ads when users are talking about sensitive topics, Markey calls into question whether the company will still use this information to personalize ads in subsequent chats.
“AI companies have a responsibility to ensure that AI chatbots do not become another digital ecosystem structured to covertly manipulate users,” Markey writes. He’s giving OpenAI, Anthropic, Google, Meta, Microsoft, Snap, and xAI until February 12th to answer questions about ads in AI chatbots and what they’re doing to protect users.