‹ BackHN Continuity

Thread

FTC is investigating OpenAI, Anthropic and other AI companies over product risks

210 points · 160 comments · dgellow

  1. Cipater · · focus · HN ↗
    They just need to hire the right lobbyists, get a meeting at the White House, agree on the terms and call it a settlement.

    And if the FTC chair is not agreeable, they will be threatened by said lobbyists and eventually be fired.

    If this sounds far-fetched, it is exactly what happened to Gail Slater, then-head of the DOJ's antitrust division and her deputies in 2025 over HPE’s bid to acquire Juniper.

    Wall Street Journal archive link: <a href="https:&#x2F;&#x2F;archive.ph&#x2F;IMer9" rel="nofollow">https:&#x2F;&#x2F;archive.ph&#x2F;IMer9

    1. paimapi · · focus · HN ↗
      there&#x27;s also regulatory capture where regulators get hired by these companies for six-to-seven figure jobs once they&#x27;re out

      <a href="https:&#x2F;&#x2F;insights.som.yale.edu&#x2F;insights&#x2F;us-government-regulators-may-be-favoring-their-future-private-sector-employers" rel="nofollow">https:&#x2F;&#x2F;insights.som.yale.edu&#x2F;insights&#x2F;us-government-regulat...

      1. wins32767 · · focus · HN ↗
        That&#x27;s not an inherently bad practice. I&#x27;d much rather there be two way flow between industry and government than have people who have never tried to run a business doing regulation or those who have never worked in government trying to comply with regulations. Strong ethics rules and culture is the way you manage the downside, not bans.
        1. salawat · · focus · HN ↗
          The hell do you work where ethics culture exists in business? I gotta be looking in the wrong places, because all I ever see is lip service granted to ethics until they threaten the bottom line, then suddenly they disappear.
          1. SpicyLemonZest · · focus · HN ↗
            The AI discourse is full of cases where companies and their employees respect ethical boundaries at a cost to their bottom line. There was a famous incident, Project Glasswing, where Anthropic announced that they needed to do a large, collaborative cybersecurity project before they could ethically sell their newest model.

            Of course, this story also points towards how people sometimes get confused. Most opportunities companies have to be unethical are complex, so if you&#x27;re not already inclined to trust a company, it&#x27;s easy to come up with reasons to distrust any claims of ethical behavior they make. Maybe they didn&#x27;t want to do the unethical behavior in the first place, maybe they&#x27;ve made up self-serving ethical standards, maybe they&#x27;re just being performatively ethical as a marketing strategy and they&#x27;ll misbehave when nobody&#x27;s looking.

            Such commentary tends to go viral, because it&#x27;s easy to propose and impossible to refute in a short comment or video.

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.