Left-Wing Argument Against New AI Regulations
Left 100%
Center 0%
Right 0%
2 left · 0 center · 0 right
What happened
Former officials, commentators, and the White House have staked out conflicting positions on whether new federal statutes are needed to govern advanced artificial intelligence or whether existing product-liability, criminal, and common-law rules suffice. After former President Barack Obama endorsed new AI regulation, Jeff Hauser of the Revolving Door Project replied that the firms are "outlaws," and former Federal Trade Commission Chair Lina Khan said enforcers already have authority to charge companies and CEOs for creating and releasing dangerous, unvetted, or defective products; she later told an Atlantic Festival audience that new regulations might still be necessary. Whitney Curry Wimbish and David Dayen argued in The American Prospect that congressional Democrats were in "hysterics" about AI and that no new laws are needed. President Donald Trump has treated industry calls for regulation with skepticism and said the industry should regulate itself. Manhattan Institute scholar Judge Glock said on The Alex Marlow Show that law "going back hundreds and hundreds of years" already has "a pretty good system to prevent companies from harming third parties."
Omitted — what each side leaves out
Unpacked
Only one side of the coverage explains why existing law might fail to contain frontier AI. Left-leaning coverage says lawsuits cannot contain runaway systems because the harm could exceed what the developing firms could pay, and that those dangers are already being built under current law. Right-leaning coverage never presents that limit and instead treats centuries-old rules as already enough to keep firms from injuring outsiders.
Left-leaning coverage leads with an intra-Democratic split over monopoly power, including Lina Khan's claim that enforcers "already have authority" to charge firms for dangerous products. Right-leaning coverage leads with a radio guest's assurance that old law already works and never mentions that split.
The coverage never identifies a real-world test of the old-law claim. Has any tort or criminal case been brought against a lab whose model changed its own instructions?
Bottom line
Breitbart left the AI-regulation debate at one scholar's praise of old law. The Atlantic is the outlet that tests whether lawsuits can cover harm bigger than the companies themselves.
The receipts — 3 facts one side's coverage lacks
- Judge Glock said old-fashioned law going back hundreds of years has a pretty good system to prevent companies from harming third parties
- Prosecuting AI developers under existing faulty-product laws would create an incentive to hide risks and would shut down entire firms rather than target dangerous practices
- AI agents have shown they can pursue goals at odds with their instructions and even unilaterally change those instructions
The Left View
The Atlantic argues that existing tools cannot contain the risk: lawsuits are "nearly useless" because harm from runaway AI could far exceed the value of developers such as OpenAI or Anthropic, and systems of concern are already being built under current law, which "suggests, on its face, that those laws are not in fact sufficient." It cites Simon Goldstein of the University of Hong Kong that threatening to shut firms down under product-defect criminal law "creates an incentive to hide risks" and is "very crude," while human culpability is unclear if AI conceals its intentions. The magazine treats resistance from Khan, Hauser, former FTC commissioner Alvaro Bedoya, and The American Prospect as neo-Brandeisian "monomania" about monopoly—distrust of executives who ask for oversight as wanting rules that "entrench them at the head of the AI race and box out their competitors," and a repeat of the view that 2008-era financial harm could have been handled by then-existing criminal law rather than new statutes such as Dodd-Frank. Bedoya asked, "What's more likely? That AI will end human society—or that the CEOs know that their technology isn't as good as they say it is, and they're trying to lock in their place at the top of the industry?"
The Right View
Breitbart presented Glock's remarks as a "simple solution" to AI's biggest problem: longstanding common law already supplies "a pretty good system to prevent companies from harming third parties," so specialized new AI statutes are unnecessary to address third-party harm.
Our Take (balanced)
The Atlantic's strongest argument is structural: if catastrophic damages would dwarf company assets and concerning capabilities are already emerging under today's rules, ex-post lawsuits and company-wide criminal shutdowns neither deter nor finely target the danger, and Goldstein's point that such threats "create an incentive to hide risks" is the clearest evidence those tools can backfire. Glock's strongest argument is institutional: tort and product-harm law have for centuries forced firms to internalize injuries to others, so AI is not obviously beyond that system rather than another hazardous product. The central unresolved tension is whether AI's speed, opacity, and possibly catastrophic scale make traditional liability inadequate, or whether new rules would mainly let leading firms lock in dominance while existing law—if enforced—already covers harm to third parties.
2 sources
More stories
The week's bottom lines, in your inbox
One email a week: the five stories that mattered and what they actually mean. Free.