Staff Reports
Judge Glock’s short interview on The Alex Marlow Show this week put a clean, old‑fashioned idea back in the ring: don’t rush to build a giant new AI bureaucracy. Use the law we already have. Glock, Director of Research and Senior Fellow at the Manhattan Institute, told host Alex Marlow that “the old‑fashioned law we’ve had…going back hundreds and hundreds of years actually does have a pretty good system to prevent companies from harming third parties.” That line is getting attention because it points to a simple answer for a big problem: more civil lawsuits, not more regulators.
What Glock is actually proposing
Glock’s fuller argument, spelled out in a City Journal essay, is straightforward. Let tort law — negligence and product‑liability suits — make firms pay when their AI harms people who are not their customers. The idea is basic economics applied to law: the party who can prevent the harm most cheaply should bear the cost. If AI labs know they will be sued for careless products or deployments, they will act more safely. No new agency. No novel licensing regime. Just courts doing what courts do.
Why this practical approach matters
This idea matters because it plays to our strengths. American courts can move faster than Congress when harms happen. Plaintiffs’ lawyers already know how to build cases. Insurers and risk managers react quickly when liability risks rise. If firms have to document safety work or face big losses, you will see safer products and smarter rollouts. It also preserves innovation. Unlike sweeping rules from Washington or Brussels, lawsuits let judges and juries sort bad actors from responsible ones, not one‑size‑fits‑all rulemakers with agendas.
Yes, there are limits — and yes, critics will scream
Let’s be honest: torts won’t solve every nightmare. Cross‑border defendants and tiny, judgment‑proof outfits can dodge responsibility. Systemic, catastrophic risks may need backstops beyond ordinary lawsuits. And of course the left will jump on every tragedy to demand new agencies and controls. But that doesn’t erase the value of using tried‑and‑true civil liability to push companies to act responsibly. The smarter path is to use courts to force firms to internalize harms while lawmakers debate bigger fixes — not to hand regulators a blank check to rewrite tech policy overnight.
What to watch next
If Glock’s view gains traction, expect more nuisance and high‑stakes suits against AI labs and platforms. Insurers will change coverage, and companies will document their safety work or face hard questions in court. But don’t forget parallel tracks: state and federal bills are still moving, and Europe is pushing an EU‑style mix of pre‑rules and product‑liability changes that will shape global firms’ choices. For conservatives who favor limited government and strong incentives, Glock’s pitch is worth taking seriously: let the law do the heavy lifting, and leave regulators out of the lab unless absolutely needed.
