A friend of mine got a message last week from a coworker — an obituary. A mutual friend had apparently died suddenly. The post spread across Facebook, shared dozens of times among friends and family, before someone noticed the deceased woman sitting at her desk, very much alive.
The Facebook group it came from is dedicated to Iowa State Cyclones content. It’s entirely AI-generated. Fake stories, convincing enough to fool people before anyone stops to verify them.
That’s where we are today. And it’s only the beginning.
This month, Anthropic — an American AI company — announced it had built a model called Mythos so capable of finding and exploiting hidden flaws in software that the company deemed it too dangerous to release publicly. Britain’s government independently tested it and confirmed it could carry out complex cyberattacks that no previous AI model had completed. Canada’s finance minister compared the threat to closing the Strait of Hormuz. A Russian state outlet called it “worse than a nuclear bomb.”
Anthropic is currently sharing Mythos only with a small group of American companies — Amazon, Apple, and Microsoft among them — to help patch the vulnerabilities it finds before adversaries can exploit them. The window is short: Anthropic estimates that other groups, including foreign ones, will develop similar capabilities within 18 months.
I understand why some Iowans are uneasy about artificial intelligence. Concerns about jobs, the economy, and misuse of the technology are real, and I don’t dismiss them. But the push to regulate, constrain, or slow AI development in the United States is a serious strategic mistake — one we may not get a second chance to correct.
The fake obituary spreading through an Iowa sports group is a small, early example of what AI-enabled deception looks like. What’s coming is the infiltration of critical infrastructure — power grids, financial systems, communications networks. Those threats will require AI-powered defenses to counter them. No human-scale response will be fast enough.
And the economic disruption will be severe regardless of what we do. Dario Amodei, CEO of Anthropic, has warned that AI could eliminate up to half of all entry-level white-collar jobs and push unemployment to between 10 and 20 percent within five years. Verizon CEO Dan Schulman has predicted unemployment will climb to 20 to 30 percent within two to five years as AI and robotics reshape the economy. These aren’t fringe predictions — they’re coming from the people building and deploying this technology. And unlike past waves of automation, there may be no “other side” to wait for, because AI’s capabilities will keep growing after the jobs are gone.
The question isn’t whether disruption is coming. It’s whether America will be holding the tools to navigate it — or scrambling to catch up to adversaries who are. This week, China’s DeepSeek released its new V4 model, claiming it rivals the best American AI systems in reasoning and autonomous capabilities. More alarming: it runs entirely on Chinese-made chips, meaning U.S. export restrictions — one of our few remaining leverage points — are being actively engineered around. Meanwhile, the White House this week accused Chinese entities of running industrial-scale operations to steal capabilities directly from American AI models. China’s AI community watched the Mythos announcement with alarm. For now, they don’t have what we have. That won’t last if we constrain ourselves while they don’t.
There is no international treaty governing AI. No shared inspections. No agreed-upon rules. Whoever builds the most powerful models gains outsize geopolitical advantages. The impact on Iowans will be catastrophic if we lose this New AI War. Iowa families deserve honest leadership on this. Limiting American AI development won’t protect workers or preserve trust — it will simply hand the future to someone else.
- Joe Monahan
Boone
