I’ve got a confession: I’ve got AI on my brain. For the past year I’ve been writing my next book, titled The Catholic Guide to Artificial Intelligence, and I’m about to turn it into the publisher (Deo Gratias!). The interweb algorithms must know this, because I’m bombarded with stories about AI every time I venture to news or social media sites. It also seems every conversation ends up focused on AI, but that might be my fault. And of course, Pope Leo just released an encyclical on the topic, which I reviewed on a recent podcast.

One thing I’ve noticed is a growing and strong anti-AI feeling among many young people. Not just, “I’m worried AI will take my job,” but “I think AI is evil and I want it to burn in the lakes of fire.” Sure, the job anxiety might be fueling the strong antagonism, but I think it’s much more than that. Young people have been living in a virtual world their whole lives and at least a certain cohort has had enough, and AI is the tipping point. They want reality, and the physical world is far more attractive to them, warts and all, than a sycophantic robot doing their work for them.

The biggest news in the AI world is the latest battle between Anthropic and the U.S. government. The conflict began a few months ago, when the Department of War blacklisted Anthropic’s AI software. As I wrote about it then:

The Trump Administration has blacklisted the AI company Anthropic because it wouldn’t play ball in allowing the Pentagon to use its software however it wanted. Trump officials say it would only be used for “lawful purposes,” but Anthropic is worried its AI could be used for massive surveillance or for fully autonomous weapons, i.e., weapons in which an AI could kill a human without any human approval. 

On Friday the Department of Commerce issued an emergency export-control directive ordering Anthropic to immediately suspend access to its latest model “Fable 5” for all foreign nationals, even including foreign nationals who worked at Anthropic developing the software.

What’s this mean? Let’s first take a step back. Fable is a stripped-down version of “Mythos,” which is Anthropic’s latest AI model that is particular used for cybersecurity. A few months ago Anthropic announced that Mythos was so powerful in testing—Mythos found serious bugs in just about every web browser and operating system in use today—that they wouldn’t release it to the public, for fear it could be used for nefarious purposes.

Fable, then, was Mythos released to the public with guardrails; it would prevent users from using it to hack into website and computers. Yet just three days after release, the federal government claimed those blocks were defective, saying researchers found ways to “jailbreak” Fable and skirt the guardrails. And while Anthropic was only ordered to keep it away from foreign nationals (including those they employed), in the end they decided to shut it down completely. Anthropic says that the government is overblowing the bugs in Fable.

It’s hard to say who to trust in this situation. On the one hand, AI company’s aren’t known for their ethical standards, but on the other hand the federal government clearly has an axe to grind with Anthropic. But regardless, it’s an example of the war that’s ramping up between AI companies and governments. How exactly should AI be regulated? Considering these systems’ powerful capabilities, it’s not unreasonable for governments to what to curtail their usage. Yet these same governments won’t think twice to abuse those same tools for their own benefit.

One thing I realized in my research into this field: there are no easy answers. Even after writing a book on the subject, I still wouldn’t classify myself as either “pro-AI” or “anti-AI.” Like the pope, I see it as a valuable tool, but one fraught with dangers. And as much as some people might like to throw all AI into a raging bonfire, we’re stuck with it. So as Catholics we’re going to have to learn to live with it and use it in a responsible way.

Hey, that’s a good idea for a book…