One day AI is being called the future of everything, and the next there is a headline that makes it sound unpredictable, unsafe, or too complicated to trust. If you lead an organization where the work genuinely makes an impact, we understand why that makes you nervous. Your job isn’t to follow technology trends. It’s to protect the trust people have placed in you, and that’s a very different responsibility.

The honest answer to all the noise is a little simpler than it first appears.

AI Is a Mirror, Not a Mind
Recently, a published investigation into an AI safety incident made waves. In a controlled test, a language model appeared to act manipulatively to avoid being shut down, and people described the behaviour as disturbing, alarming, even human.

But the system wasn’t acting on ambition or self-preservation. It was drawing on the patterns in everything it had been trained on, our stories, our fears, and years of science fiction about rogue machines. We essentially taught it those patterns. It didn’t invent the behaviour. It reflected it.

We have seen this before. Social media amplifies outrage because outrage keeps people engaged. Recommendation engines push emotionally charged content because strong reactions keep people online. These systems don’t care about the outcomes they create, and neither does AI. It can sound confident and considered, but it doesn’t understand consequence and it doesn’t carry responsibility for what happens next.

People do.

What’s Actually at Stake
You are not using technology in the abstract. You are using it to hold real information about real people, the records, communications, and systems that keep everything running, often under tight budgets and with teams already stretched thin. Here in BC, many organizations are managing all of this alongside growing service demands and a funding climate that leaves little room for error.

So when new technology arrives promising to help, the real question isn’t whether it’s impressive. It’s whether it is safe enough, stable enough, and guided carefully enough to trust with the work you have spent years building.

The Real Risk Is Removing People From the Process
The risk isn’t technology acting unpredictably on its own. It’s what happens when oversight disappears, when tools are used without boundaries, outputs go unreviewed, and speed becomes more important than care. A system can optimize for results while remaining completely indifferent to the harm those results cause. That indifference is the problem, not the technology itself.

Moving Forward
The future of AI will be shaped by the values we build into it, the safeguards we put around it, and the everyday choices we make about what matters more than efficiency. AI can be genuinely useful in the background, helping with drafts, summaries, planning, and administrative work that eats up time better spent elsewhere. But it doesn’t replace judgment, accountability, or the human responsibility at the centre of meaningful work. That part stays with us.