The Economist likens AI to the Wild West frontier in this slightly tongue-in-cheek article. But the bit about barbed wire is right.
The article calls for better AI guardrails because, like the Wild West of old, “the [AI] frontier is lawless”.
The recent breakouts (or “loss-of-control” in PR-speak) by OpenAi and Anthropic models are good examples of bad things happening.
And when bad things happen, there’s money to be made. The share prices of major cyber-security firms have roughly doubled this year.
When it came to the taming of the Wild West, apparently Alan Greenspan thought barbed wire was important because it “boosted productivity by allowing the settlers…to safeguard their properties”.
AI buzzword of the year 2026 is “harness”*. That’s the system that surrounds you LLMs to keep them and your agents on the straight and narrow. The article suggests that your harness should be made of barbed wire. That’s good advice.
* According to me. If you want to know more about harnesses (yes, they are real) check out our AI hero Nate B. Jones. He has some really videos and articles on the topic.
