FTC takes aim at AI bias, warns ‘ideological objectives’ could violate federal law
Technology companies that build chatbots to push ideological agendas could be violating a federal law that prohibits “unfair or deceptive” conduct, the Federal Trade Commission said Wednesday as it took aim at a growing “patchwork” of state AI regulations.
Tech companies could manipulate AI-powered chatbots to deliver politically biased responses in ways that are “contrary to reasonable consumer expectations for objectivity and accuracy,” the commission said on Wednesday. Doing so, officials said, could violate federal rules against deceptive business practices.
“Customers have a reasonable expectation that AI systems aim to give truthful and accurate outputs,” the commission wrote in a proposed policy statement.
Through a public comment period, the commission is seeking feedback on ways people believe they are being subtly steered to information that aligns with the “undisclosed ideological objectives” of tech companies.
“Customers have no basis to believe that AI systems aim to produce outputs that are distorted by undisclosed ideological objectives,” the proposed policy statement notes. “Nonetheless, an AI company might be tempted to alter or steer the output of its systems contrary to consumers’ reasonable expectations for various reasons, including attempted compliance with a state law.”
“The FTC wants to hear from businesses and consumers about their experiences and concerns regarding the subversion of AI systems for ideological ends,” Chairman Andrew N. Ferguson said in a statement. “This crucial input will help the Commission formulate a final policy that advances President Donald Trump’s goal of expanding America’s global dominance in artificial intelligence.”
The commission’s policy proposal notes a reasonable consumer would not expect AI systems to provide responses that are “clearly and unequivocally not protected by the First Amendment,” like child pornography, and that federal officials do not plan to “prohibit companies from imposing limits” that prevent cybersecurity attacks. While the FTC proposal has its own First Amendment implications, the commission said state laws that restrict “truthful speech because it might lead another person to commit” discrimination violate free speech rights.
Why is the FTC taking aim at state AI laws?
The move isn’t just about tech company bias. It’s part of a broader national battle over who gets to regulate AI-generated outputs — and the commission specifically calls out state efforts to prevent AI-enabled discrimination. It follows a December executive order by President Donald Trump directing the commission to address the legality of state AI regulations. In January, Trump signed another executive order that took aim at “Woke AI” chatbots and demanded they produce responses free of ideological biases.
Federal officials have especially targeted new AI regulations in Colorado. The commission alleges a Colorado law coerces tech companies to alter chatbot outputs “to comply with and advance the state’s ideological objectives.”
Earlier this year, Colorado overhauled its first-in-the-nation AI law, replacing strict anti-discrimination risk assessments with disclosure rules that mandate transparency about high-stakes decisions regarding education, healthcare and employment.
The FTC announcement Wednesday suggests the government is examining AI regulations through a lens of consumer protection — and places the federal government in a position to define truth and bias in the AI age.
A growing body of research suggests that leading AI chatbots have biases. Most of them skew to the left, according to an investigation published this week by The Washington Post. The worst offender was OpenAI’s ChatGPT, the Post found, while Google’s Gemini took a both-sides approach that balanced liberal and conservative positions in 90% of chat responses.
Round out your reading
- America at 250: Straight Arrow’s week-long look at the pressures testing modern America.
- Ketamine is booming in America. The reasons go beyond drugs.
- Cluster of AI news sites suddenly goes dark after human reporter starts asking questions.
- A 79-year-old, obese, well-connected man gained exclusive access to a new weight loss drug. Is it Trump?
- Newsom pushes national wealth tax after opposing California billionaire tax.
Ella Rae Greene, Editor In Chief
