OpenAI's ChatGPT Ad Plans Spark Privacy Concerns, Prompt Researcher Exit

Alanbatnews -

OpenAI's plan to introduce advertising within ChatGPT has ignited a debate over user privacy and ethical considerations, leading to the resignation of a researcher and raising concerns among competitors.

Zoe Hitzig, formerly a researcher at OpenAI, stepped down from her position in protest of the company's decision to incorporate advertisements into ChatGPT, drawing parallels to the advertising-driven models of Facebook and Meta.

Hitzig voiced her concerns in a detailed account published by The New York Times, explaining that her initial motivation for joining OpenAI was to contribute to the development of AI models and establish early safety policies.

However, she says OpenAI has shifted its focus away from these core principles, abandoning the pursuit of answers to the very questions that drew her to the company.

Hitzig's apprehension is shared by other experts, including those at Anthropic, a rival AI developer. Anthropic has publicly stated it has no plans to introduce advertising into its AI tools, according to a CNBC report.

Demis Hassabis, CEO of Google DeepMind, also expressed surprise at OpenAI's move towards advertising, according to TechCrunch. The central concern revolves around the potential misuse of sensitive user data.

Hitzig emphasizes that her objection is not to advertising per se, given the substantial costs associated with AI development and operation. Instead, she highlights the ethical dilemma of leveraging tools that possess sensitive user information for advertising purposes.

“People are talking to ChatGPT about the most sensitive and private aspects of their lives, with the belief that the tool has no ulterior motives,” Hitzig stated.

She notes that users confide in ChatGPT about a range of personal matters, including religious beliefs, personal and psychological struggles, professional challenges, and medical anxieties, potentially including information about others.

While OpenAI intends to adhere to safety and privacy standards in its advertising practices, Hitzig fears that these safeguards may erode over time as users become accustomed to the presence of ads within the platform.

Hitzig draws a comparison between OpenAI's current advertising strategy and the early practices of Facebook, which initially pledged to protect user data and prevent its use in advertising, a promise that has since been abandoned.

She raises a fundamental question about OpenAI's ability to create a framework that does not treat user data as a commodity to be sold to advertisers.

Anthropic has distinguished itself by explicitly rejecting advertising, even running a Super Bowl advertisement to emphasize its commitment to ad-free AI tools. The company's revenue model relies on partnerships with major corporations, providing AI solutions and user subscriptions, according to CNBC.

Hassabis, while not explicitly ruling out advertising in Google DeepMind's tools, indicated that his team is approaching the matter with extreme caution, according to TechCrunch.

He added that Google's leadership has not mandated or pressured the team to incorporate advertising into its AI models, despite the fact that advertising is a significant component of Google's revenue generation strategy.

Hassabis differentiates between search engines and chatbots, arguing that while advertising works well with traditional search, the “digital assistant” model depends on establishing a relationship of trust with the user to provide personalized assistance. He questions how advertising can be integrated into this model without undermining that trust.

ChatGPT boasts over 800 million monthly active users. While OpenAI offers a range of paid services, the majority of users rely on the free version of the tool, according to the Associated Press.

Despite an initial valuation of $500 billion, OpenAI is reportedly spending more money than it is earning, necessitating a new approach to achieve profitability.

OpenAI maintains that user data from conversations will not directly influence the advertisements displayed. However, The Conversation suggests that self-imposed restrictions by companies often fail in the long run, echoing Hitzig's concerns and the motivation behind her resignation.