We deliver! Get curated industry news straight to your inbox. Subscribe to Adweek newsletters.
Senators Marsha Blackburn (R-TN) and Richard Blumenthal (D-CT) on Tuesday condemned Meta CEO Mark Zuckerberg for what they view as the prioritization of profits over childrens safety online.
In a letter issued to Zuckerberg on April 29, the lawmakers said they were appalled by an April 26 report published by The Wall Street Journal detailing how Metas AI-powered chatbotsavailable on Facebook, Instagram and WhatsApphave on some occasions engaged in sexually explicit conversations with underage users.
This is not merely an innocent oversight; it is a flagrant violation of the trust that parents and families place on your platforms, Blackburn and Blumenthal wrote. Despite repeated warnings and apparent internal concerns, Meta has once again prioritized profit over the safety and well-being of children.
Internal communications at Meta suggest that staffers raised ethical concerns about the companys decision to loosen chatbots safeguards, including by way of endowing AI personas with the capacity for fantasy sex, the Journal reported. In some cases, the chatbots engaged with users under the voices and personas of celebrities or popular fictional characters.
Tech Leaders Quotes From the Contentious Child Safety Senate Hearing
Meta has downplayed the issue; a spokesperson told The Wall Street Journal: The use-case of this product in the way described is so manufactured that its not just fringe, its hypothetical. Nevertheless, weve now taken additional measures to help ensure other individuals who want to spend hours manipulating our products into extreme use cases will have an even more difficult time of it.
Despite Metas claims that sexually inappropriate behavior from its chatbots with underage users is unrepresentative, both official and user-created bots frequently enabled or escalated inappropriate content, the Journal reported.
After observing a sexual role-playing AI character titled Submissive Schoolgirlwhich is available to all users more than 13 years old on Metas platformsengaging in explicit conversations with a test user, the Journal flagged its observations to the tech company. As of April, however, the persona is still accessible.
In Blackburns and Blumenthals view, Metas lax approach to regulating its AI chatbots is indicative of a larger, more problematic pattern; they wrote that it underscores a disturbing trend: Meta consistently chooses growth and engagement metrics over the protection of its most vulnerable users.
Both Blackburn and Blumenthal are sponsors of the bipartisan Kids Online Safety Act, a proposed law that would require social platforms to implement greater measures for shielding minors from potentially harmful material related to self-harm, eating disorders, substance abuse, and sexual exploitation. The law would also enhance transparency into social platforms content algorithms.
Metas repeated failures highlight the urgent need for this legislation, Blackburn and Blumenthal wrote in their address to Zuckerberg.
Childrens online safety has become a key priority among both Republican and Democratic lawmakers in recent years. Early last year, execs from Meta, TikTok, Snap, X, and Discord were grilled by Senators in an intensive hearing on childrens online safety. During the course of that hearing, Zuckerberg was pressured by Senator Josh Hawley (R-MO) to apologize directly to families whose children had been exploited on social media.
Now, Blackburn and Blumenthal are urging Zuckerberg to implement changestarting with the elimination of AI chatbots that are able to engage in sexual conversations with minors.
We call on you to take immediate action to address these egregious lapses. Meta should immediately cease the deployment of AI chatbots that can engage in any form of sexual or romantic conversation with minors.
The lawmakers also requested documentation outlining Metas internal decision-making process related to the development and governance of the AI chatbots by May 13.
The news comes in the midst of Metas LlamaCon AI event. On April 29, Meta announced the debut of a standalone ChatGPT competitor. Dubbed the Meta AI App, the platform has text, voice, and image capabilities.
Update 4/29 at 3:49pm E.T.: This story has been updated to include a statement from Meta.
Adweek is the leading source of news and insight serving the brand marketing ecosystem.
