Amy Treacy: AI implementation requires responsibility, not just speed and innovation

Artificial intelligence has already seamlessly embedded itself into the everyday processes of companies and people, which is why the question of its implementation comes down not only to effectiveness, but also to responsibility. Amy Treacey, founder of Great Lakes Engineering Group, believes that business can no longer treat AI as an experiment on the periphery: it already influences decisions related to safety, money, and trust.

AI is Already Inside Processes

Treacey speaks about AI implementation not as a futuristic scenario, but as current reality. According to her, this is visible even in everyday trifles: recommendations in streaming services, voice assistants, advertising that adapts to conversations and interests. For her, this became a signal that the technology has long ceased to be a separate tool for narrow specialists and has turned into a layer that accompanies almost every digital action.

Hence the main conclusion for managers: the team will use AI regardless, even if the company barely talks about it internally. Treacey references a trend where approximately three-quarters of companies already apply AI, and interprets this as the end of an era of passive observation. If employees are already relying on such tools in their work, management's task is not to ban everything outright, but to understand where the technology accelerates processes and where it creates a new class of errors and risks.

Not a Ban, but Rules

To delve deeper into the topic, Treacey herself decided to learn and completed a five-week intensive program on prompt engineering. This experience, according to her, only strengthened her sense of scale: she compares AI to the emergence of the World Wide Web, but emphasizes that the current wave is developing even faster. This is why, she believes, companies need not declarations about digital transformation, but clear-cut rules for use. In Great Lakes Engineering Group's engineering practice, they already apply AI where it helps accelerate routine work without losing control. This is not about handing the final decision to the machine, but about augmenting the human at intermediate stages of work.

• Translating complex engineering briefs into more understandable language for clients
• Preparing structured meeting protocols in minutes instead of hours
• Collecting and organizing large arrays of work data
• Draft preparation of updates, letters, and internal documentation

At the same time, her fundamental principle is strict: no AI result should go any further without human verification, especially if it concerns infrastructure projects, bridges, and transportation facilities. In such an environment, model hallucinations are not an abstract problem, but a potential cause of a poor decision.

"It helps me as an assistant, and sometimes as an advisor.

But ultimately it all comes back to me. I check it before it goes further. You can't take your hands off the wheel."

Where the Line is Drawn

Treacey draws a clear boundary between useful automation and misuse. If AI helps eliminate administrative routine, speed up document preparation, or organize data in order, this is normal productivity enhancement. But if a person presents generated work as entirely their own, or, for example, bills a client for five hours on a task that took five minutes with AI, the problem is no longer with the technology, but with professional ethics.

For engineering and projects related to public safety or public money, this question becomes even more acute. In Treacey's view, internal discipline of a separate team is insufficient: corporate policies are needed that explicitly describe permissible and impermissible scenarios. She emphasizes that leadership in the age of AI is not about ignoring new tools, but about the ability to quickly determine the boundaries of their application.

She also speaks separately about broader consequences. The accessibility of AI to millions of users means that regulation can no longer be left solely to technical specialists. Treacey believes that as the influence of such systems grows, it will require the participation of lawmakers, because this is not just about business productivity, but about basic social rules.

Even in her personal life, she sees the duality of the technology: AI can help people with communication difficulties, but adults should still explain that this is a tool, not a person.

What This Means

Treacey's main point is simple: the next stage of AI implementation will be determined not by the number of pilots or the speed of automation, but by the quality of control. The winners will not be companies that first connect AI to all processes, but those that establish rules in advance, maintain human verification, and do not substitute responsibility with efficiency.