AI Briefing: The FTC is leaning into ‘tech censorship’

“Tech firms should not be bullying their users,” Andrew Ferguson, the FTC’s chair appointed by President Donald Trump, said in a statement. “This inquiry will help the FTC better understand how these firms may have violated the law by silencing and intimidating Americans for speaking their minds.”

The FTC’s document requesting public comment includes topics to learn whether online platforms took action against users and whether users felt actions were related to speech on or off the platform. The agency also asks about platforms’ policies related to content moderation and appeals processes, and how the decisions impacted users. Other questions ask about platforms’ motivations for various policies and whether actions were a result of a lack of competition or affected competition.

Some of the questions seem to echo allegations made in X’s lawsuit against the World Federation of Advertisers. The suit alleges that WFA and various advertisers conspired to boycott ad spending on the platform formerly known as Twitter over brand safety concerns after Elon Musk acquired it.

Examples of FTC questions [synthesized]

• Did platforms implement similar policies compared to other platforms or take comparable adverse actions?
• Did platforms coordinate on policies and adverse actions directly or indirectly?
• To what extent have platforms financially supported or partnered with organizations that advocate for or enable censorship (e.g., through advertising boycotts)?
• Did platforms gain market dominance under permissive content policies and then change them after achieving market power?
• Have platforms used their dominance in other markets to suppress competitors with different moderation practices (e.g. search engines and app stores)?

According to the FTC’s website, more than 300 comments were submitted in the first 24 hours. Many of the comments submitted have been anonymous.

In response to the agency’s inquiry, the tech-funded lobbying group Chamber Of Progress issued a statement that noted allegations of tech censorship don’t mention conservative publishers and users often earn higher engagement than liberal counterparts. (The press release was published with the headline “FTC Chair Rides MAGA ‘Tech Censorship’ Hobby Horse.”)

“By attacking content moderation, the FTC undermines its pro-competition mission. Moderation lets platforms offer distinct user experiences, and restricting it risks forcing a one-size-fits-all internet,” according to the Chamber Of Progress’s statement.

Since Ferguson replaced Lina Khan as FTC chair last month, there’s been speculation about how tough Trump’s pick might be on Big Tech. Some legal experts expect Ferguson to take a more laissez-faire approach to tech regulation, while others predict the president or other political winds could sway his priorities.

Conservatives’ attacks on what they see as censorship aren’t anything new. Last year, the U.S. Supreme Court issued unanimous rulings to block state laws in Texas and Florida that sought to regulate social platforms’ online moderation. Both laws were initially passed in 2021 after Trump was kicked off Twitter and Facebook over his posts related to the Jan. 6 attack on the U.S. Capitol.

The high court ruled platforms’ policies are protected by the First Amendment, not in violation of it. In its opinion, the court decided companies are allowed to curate and edit third-party speech while a state can’t “advance some points of view by burdening the expression of others.”

Another AI unicorn

In other news, this week saw (yet another) AI unicorn anointed, with Hightouch announcing it has raised $80 million — funds it will use to further its vision for agentic marketing and its latest product AI Decisioning.

The company claims that “the old way of marketing wasn’t working, so we invented a new one.” Sapphire Ventures, NVC, Bain Capital Ventures, ICONIQ Growth, Y Combinator, Afore Capital and Amplify Partners all signed off on those claims, collectively valuing Hightouch at $1.2 billion.

“Instead of pre-planning every experience using rigid rules and timing, marketers can simply assign a goal like ‘drive customers who buy in-store to download our app and order ahead’ and then watch as AI agents use approved content to deliver personalized communications,” reads the announcement from Hightouch’s C-suite.

CriteoPrivateAd

Meanwhile, Criteo — the storied ad tech company that pioneered ad retargeting — is marking the tenure of its new CEO, Michael Komasinkski, with an open-source project from its AI Lab that will explore privacy-enhancing technologies, or “PETs,” as the industry collectively terms them.

To this end, the company is open-sourcing CriteoPrivateAd — “the largest real-world anonymized bidding dataset” — in a scheme it hopes will help steer its strategic decisions as the industry (slowly) prepares for the retirement of third-party cookies.

CriteoPrivateAd aims to train and evaluate privacy-preserving machine learning models for digital advertising, including RTB strategies, based on the real-world bidding log. It also mimics the constraints of browser-based APIs like Chrome’s Privacy Sandbox — an area where the company has been a leading voice — to help researchers model ad bidding under realistic private limitations.

“We are releasing the largest, in terms of number of features, bidding dataset specifically built in alignment with the design of major browser vendors proposals such as Chrome Privacy Sandbox,” reads the introduction to a paper accompanying the launch of CriteoPrivateAd. “We ensured that this dataset while being anonymized, is able to provide offline results close to production performance of ad tech companies, including Criteo — making it a relevant ground truth to design private advertising systems.”

Boardroom battles: Marketing vs. finance

Despite all the hype, a recent survey from Gartner reveals that 27% of CMOs claim their marketing organization has limited or no adoption of generative AI in marketing campaigns.

“High-performing marketing organizations are integrating GenAI at a faster pace than their peers,” reads the report, which categorizes “high-performing” marketing operations as “1.3x more successful in overachieving year-on-year profit growth.”

Researchers have found that the 27% who have yet to use such technologies are most likely hindered by their colleagues in the finance department — who knew?!

The study continues, “31% of marketing leaders say that finance is the function that most impedes successful execution of marketing campaigns, followed by executive leadership (26%) and sales (26%).”

However, marketers’ boardroom detractors could also point to a separate study from eMarketer (also published last week) that claimed 64% of U.S. adults are uncomfortable with AI-generated ads. Despite this, 48% of respondents still think AI will be a popular way for brands to advertise products in the next few years.

Other AI news from the week

• Google announced new creative updates for advertisers within Google Ads. Updates include ways to create AI-generated images of young adults, recommendations for asset audiences and new asset testing for Performance Max campaigns.
• Perplexity debuted its own version of DeepSeek R1 (“R1-1776”), which it claims to be free of Chinese censorship.
• Mira Murati, the former CTO of OpenAI, launched a rival AI startup called Thinking Machines to help make AI systems more understandable and customizable. She’s already poached more than two dozen researchers from OpenAi and other rivals to join her.
• Elon Musk’s AI startup xAI released Grok-3, the next version of its ChatGPT rival.