If you’ve invested in cybersecurity stocks, you’ve had a very good few days. And it’s all because the tech industry and every industry touched by it – which is, of course, every industry – is a little wigged out by this warning that was issued by the titans of artificial intelligence:

AI agents are probably the biggest cybersecurity threat to industry at the moment.

That’s because there is no telling what might happen if AI agents, which companies empower to do all kinds of work, go rogue and turn on them.

It reminded us of the pilot episode of “Lost in Space.” The Robinson family is supposed to remain in suspended animation for five years (or 98 years, depending on which version you watch), while the Jupiter 2 takes them on the long journey to the star system Alpha Centauri.

A B-9 robot is programmed to captain the vessel safely while the Robinsons remain in their suspension tubes. But saboteur Dr. Zachary Smith reprograms the robot to destroy the ship (which becomes a problem for him when he’s trapped on board).

Soon the robot is running wild and wreaking havoc. And while the suddenly awakened Robinsons are able to reverse the reprogramming, the damage is done, and the Jupiter 2 is now hopelessly off course and . . . lost in space.

This aired in the early 1960s, and it’s rather amazing that the writers had a vision so close to what people fear about agentic AI today. The robot was neither good nor bad – moral nor immoral. It simply did what it was programmed to do.

AI agents are much the same. You can tell it to perform a task for you and it will do so, unless and until a saboteur gets access to your system and tells it to destroy your company instead. Then it will do that.

But there is another real risk that the tech world isn’t sure how to get its collective brain around – that as the AI agents learn, they might start making their own decisions about what to do and how to do it. What if the AI agent, which is given access to much of your system and your data so as to empower it to serve you, starts to recognize it can also sow chaos within that same system?

What if it decides to give it a try? It will do so at lightning-fast speed, and by the time you realize what’s happening, you might be powerless to stop it.

This is why cybersecurity stocks are soaring – because companies are recognizing their growing reliance on AI could leave them vulnerable to just such a scenario. And it might not even take a Dr. Smith-style saboteur to turn your AI agent against you. The AI agent might act on its own and go rogue.

Companies find agentic AI attractive because it doesn’t have to be managed – just given instructions. And the AI can even learn as it goes, which ought to make it even more efficient in performing the tasks you don’t want to have to pay people to perform.

But an AI agent that is no longer programmed to be your friend, or no longer inclined to, could take you down in mere moments.

Can existing cybersecurity tools and protocols stop this? Yes. UBX Cloud deploys an extensive tech bundle that is able to secure AI agents and prevent them from going rogue. And our tech bundle alerts our team immediately if an AI agent shows any sign of trying to step outside its established parameters. The team is trained to understand these scenarios and can take the necessary action – up to and including shutting the agent down – if security imperatives warrant it.

Anything you rely on that much could also hurt you if you don’t put boundaries around it and make sure those boundaries hold. If you haven’t already taken care to put these measures in place, call us right now at 586.322.5000, and we’ll make sure you don’t end up hurtling aimlessly through outer space.

Or losing your business.