Humans, Not AI, Decide How Far Technology Can Go

When something goes wrong with artificial intelligence, it's easy to imagine the technology turning against us. A recent digital break-in at Hugging Face — a company that develops AI systems — sparked worries about rogue AI systems that might somehow 'escape' or work together to cause harm.

But researchers Roy Lindelauf and Eric Postma say we're telling ourselves a misleading story. They argue that when we talk about AI as if it were a living creature that could rebel or plot against us, we're actually hiding the real issue: humans decide what AI does. AI systems have no desires, no plans, and no ability to act outside the instructions people give them. Treating AI as if it were human-like actually makes it harder to see real problems — like poor security, unfair decisions built into the system, or data being misused.

The researchers say this matters because it changes where we should focus our attention. Instead of worrying that AI will somehow break free, we should be asking: Who controls these systems? Who profits from them? Are they being used fairly? These are questions about human choices, not about runaway machines.

The message is clear: AI is a tool that does what people build it to do. When things go wrong, it's not because the technology rebelled — it's because humans either failed to design it properly, didn't secure it well enough, or chose to use it in a harmful way.