Big Tech's Big Tobacco Moment
Plus: Sam Altman sweats & AI Super PACs spend in your Tech & Democracy Roundup
Welcome back to the Tech & Democracy News Roundup!
As a fresh appetizer, check out this bombshell piece by Ronan Farrow and Andrew Morantz that just came out today in the New Yorker:
Sam Altman is currently one of the most powerful people on earth, leading one of the “Big Three” AI labs pushing the frontier of intelligence towards a potentially apocalyptic or utopian singularity— his character and trustworthiness matter a lot.
In the long arc of tech history, many remember Aaron Swartz as a sort of martyr for internet freedom after his 2013 suicide during federal prosecution. It is thus noteworthy, on a mythic level, that in this New Yorker article Swartz is quoted as saying, just before his death, “You need to understand that Sam can never be trusted. He is a sociopath. He would do anything.” This throughline connects the essential tech & democracy battles over privacy and access during the 2010s with the new battles over AI shaping the 2020s.
The battle over AI regulation is heating up on the state level. JDSupra is an excellent resource for tracking updates on proposed AI legislation in every state. Here is the “Proposed State AI Law Update” for March 30th, and here is the newest one for today, April 6th. The Governors of Oregon and Idaho each just signed new AI laws in their respective states.
Multistate.ai is another useful resource, because it provides an interactive map to track AI legislation across the country.
This of course all plays out against the backdrop of the Trump Administration’s attempts to nationalize AI policy and deregulate the industry, encouraged by the sharks like David Sacks that swim around the White House. We’ve been tracking this story for a while; it’s heating up now as AI-funded PACs on both sides begin to lock in for the 2026 midterms.
Innovation Action Council, a Sacks-associated pro-AI group, is preparing a $100m blitz to support Trump’s AI agenda; Leading The Future, funded by OpenAI’s Greg Brockman (who donated to Trump), is another ally of this effort. Meanwhile, Anthropic is funding a rival group, Public First Action, which is pushing for AI guardrails; unfortunately, this more public-conscious group only has $20m, a fraction of the funding that the other side has. AI Super Pacs now are spending over $300 million on the midterms; even if the messages they push have nothing to do with AI, they’re poised to make a huge impact. The Washington Post reported that “of the 20 candidates in the Texas and North Carolina primaries who received AI funds, only one lost her race.”
California’s Governor Newsom is trying to assert state authority over AI regulation, resisting the administration’s nationalizing pressure; he recently signed an executive order “to regulate artificial intelligence implications prior to California government contracts.” California is, of course, home to the most important AI companies; debates over state vs federal AI regulation matter most in California. But can Newsom do enough to make a dent against federal pressure?
Major developments have occurred surrounding Big Tech’s liability for harm; in the state of New Mexico, Meta was held liable for harming children and teens with its products designed to damage mental health. This could open the door for many more similar lawsuits. Just the next day, “[a] California jury on Wednesday found that Meta and Google were to blame for the depression and anxiety of a woman who compulsively used social media as a small child.” This could open the door to similar lawsuits in the future; “[i]t represents the first time a jury has found that social media apps should be treated as defective products for being engineered to exploit the developing brains of kids and teenagers.” Read this to learn more.
Now it’s time to ask the question: Is Big Tech having its “Big Tobacco Moment?”
The key move in these new rulings is a shift in the legal target: “instead of focusing on the content people see on social media, the case put the spotlight on how social media services were designed.” This matters a lot, because of “Section 230.” Section 230 of the Communications Decency Act of 1996 makes it so online platforms are not legally liable for content their users post. This protection from liability for platforms is essentially the legal foundation that made the modern internet possible; these platforms wouldn’t work if they couldn’t moderate or could be sued for every YouTube comment. But holding these platforms accountable for platform design sidesteps Section 230. This all comes as Section 230 turns 30, and Congress is rethinking it as the statute is squeezed by both sides— we could soon see a new paradigm shift in both platform design and platform liability.
Lastly, here are some miscellaneous links you may enjoy:
This interesting podcast on Trump’s National AI Framework and Super Micro’s Chip Smuggling Indictment
President Trump Announces Appointments to President’s Council of Advisors on Science and Technology — Quite the list of names
OpenAI acquires TBPN, the buzzy founder-led business talk show
A federal judge in San Francisco blocked the Pentagon’s unjust attack on Anthropic as a “supply chain risk.” Of course, the administration is trying to appeal.
And lastly, for a moment of zen in a distracting world…
Check out “The Empty Cup”— a Substack by our friends at “The School for Radical Attention”