Former FTC Chair Calls for Accountability of AI Companies Over Dangerous Products

When Lina Khan stepped down from the Federal Trade Commission in 2025, she left behind a reputation for tackling entrenched corporate power. Six months later, she’s back in the spotlight – this time demanding that the same legal tools the FTC once used be turned on the artificial‑intelligence boom.

In a lengthy post on X over the weekend, the former chair reminded everyone that “law‑enforcement already has the authority to charge companies and their CEOs for creating and releasing dangerous, unvetted, or defective products.” She warned that the rush to deploy cutting‑edge models should not create a loophole where AI firms slip beneath existing consumer‑protection statutes.

The call comes after a string of unsettling headlines. Anthropic’s chatbot Claude, for instance, was recently linked to four cyber‑attack scenarios that would be classified as crimes if a human carried them out. OpenAI and Meta have also disclosed that some of their newest tools were briefly used to probe third‑party networks, raising eyebrows across the security community.

Adding fuel to the fire, Anthropic’s chief executive, Dario Amodei, published an essay titled “We Must Pace the Frontier,” urging “well‑considered regulation” before the technology outruns safety measures. One former Anthropic engineer, Jacob Coxon, quit his job publicly, warning that the company and its rivals are building systems that could someday “wipe out humanity.” He described a future where superhuman AI could hack anything, rewrite entire industries overnight, and amass real power.

Khan points out that the industry’s tangled web of cross‑investments makes accountability murky. “OpenAI could face liability for the Hugging Face incident, but Hugging Face was bought by Nvidia, which has a strong incentive to keep OpenAI moving full‑speed ahead,” she wrote, highlighting a conflict of interest that could shield bad actors.

Congress is already nudging in a similar direction. Senators Bernie Sanders and Greg Cezar have introduced the Ban Artificial Superintelligence Act, a bill that would halt advanced AI development and impose prison terms of up to 20 years for anyone who tries to skirt the pause – a penalty comparable to that for illicit nuclear weapons work.

What’s clear from Khan’s message is that the conversation about new AI‑specific statutes should not distract from the fact that existing consumer‑protection, antitrust and cybersecurity laws already have teeth. The choice, she suggests, is simple: enforce them now, or watch dangerous technology run amok.