The Hawley-Blumenthal AI Bill Threatens National Security

By James Erwin

Last night, News Nation hosted the fourth Republican presidential debate in Tuscaloosa, Alabama. Though these debates seem to matter little in the presidential horserace, with no one coming close to Former President Donald Trump, it was refreshing when the debate addressed the issue of Taiwan, with all four candidates expressing support for deterring a potential Chinese invasion. Unfortunately, Senators Josh Hawley (R-Mo.) and Richard Blumenthal (D-Conn.) have introduced legislation that will undermine any deterrence of the People’s Liberation Army (PLA).

Sen. Hawley has long harbored a vendetta against popular social media platforms for their censorship of conservatives. Revelations that this censorship has come primarily at the behest of government agencies have done nothing to convince him otherwise, and he continues to believe that the best path forward is to repeal Section 230 of the Communications Act, which shields tech companies from liability for what their users post. Blumenthal, for his part, takes the general progressive line that “big is bad,” regardless of conduct or competition.

As we have argued elsewhere on this site, the economic case for Section 230 is strong. The liability shield has enabled permissionless innovation in digital communications and created immense wealth for the United States. Morally and constitutionally speaking, it has enabled free speech to flourish online without the old media gatekeepers controlling what can be expressed. News outlets have grown up completely on the internet, introducing more diverse voices and competition into the media marketplace (largely to the benefit of conservatives, if Hawley still considers himself one).

The bill the senators will attempt to pass by unanimous consent today (and which will be blocked by their Commerce Committee colleagues) would carve out AI developers from 230 liability protections. This means that companies can be held liable for anything their digital tools are used for. Furthermore, any company that uses AI in the products is also captured by the bill, meaning America’s entire tech industry will lose Section 230’s protections. Here is where the national security threat becomes clear: this bill would obliterate AI development in the United States.

For AI to be useful, it must be at least powerful enough to write statements that could be defamatory, as our friends at the R Street Institute have demonstrated. With this hard threshold imposed on developers, it no longer becomes profitable, useful, or even interesting to continue development. The industry will dry up here and move to the UAE or China itself. With the U.S. military currently in an arms race against the PLA on AI development, crushing our private sector innovation will ensure we lose. If we lose the arms race, we will lose the next generation of weapons systems to our rival. At that point, the U.S. Military will no longer be much of a threat to the PLA, and we will not be able to deter an invasion. Indeed, we might hasten it.

In short, passing this bill is not only terrible for our economy, but it would also make a Chinese invasion of Taiwan even more likely. We are not given to hyperbole on this website, but these are the stakes. The Senate should not only block the bill today, but ensure it never comes close to a vote in the future.