The Principles of Ethical AI Leadership

The world’s feelings towards AI has evolved almost as fast as AI itself in the last three years. What started as something most people viewed as somewhat harmless and mostly silly has become a hotbed discussion of ethics, morality, and politics. 

Data centers – for good reason – are viewed as an existential and environmental threat, and the debate over their existence is embroiling dozens of town halls across the country.

It seems impossible to imagine middle ground in this sort of climate, but surely one can be found. Despite the feelings of the public – and potentially the employees who might be reticent to say something – is it actually possible for businesses and the leaders within them to use AI for good purposes and quality results?

AI and the 30% Rule

The evidence is often obvious; articles written by AI just feel stuffy. Like someone who learned English as a second language and only knows formal communication instead of casual colloquialisms. AI doesn’t understand how to write with sarcasm or the use of idioms, and most readers can still detect when an article just doesn’t read like a person.

And yet, no one can deny AI’s growing usage and prevalence in our everyday lives. Which is why the 70/30 rule feels more like a guideline.

The rule basically says that AI should handle 30% of a task, while the remaining 70% is handled by a human. 

Most of the reasons that corporate leaders support AI involve efficiency. They believe integrating it in their business will allow people to accomplish more in less time. And so far, results of studies are mixed on whether this is what’s happening.

It’s not about AI’s ability to assist in tasks; it’s about AI removing a person’s ability to learn and do a task on their own. The implication here is that if people aren’t able to do a task without AI’s help, then we’ll get to a point where AI is doing the entire task, leaving the 70/30 rule in the dust.

AI Infiltrates the Internet

Coders – software engineers, computer programmers, and developers – were some of the first to implement AI into their everyday lives at work. But in this industry, reliance on AI might hinder someone’s ability to complete a task, rather than aid it. 

One research study involving software developers asked participants not to use AI for assistance building a Python script. Despite that request, 35% of them did. And even when reminded not to use AI, 25% did so again.

Basically, humans have to spend time correcting AI’s mistakes. But, should use continue to grow, humans may not even know or remember how to fix those mistakes. However, there’s a strange dichotomy between a software developer’s usage of AI, and their opinion on the use of AI.

Sonar Source, a company that specializes in reviewing automated code, surveyed developers on their AI habits. It found that 42% of today’s internet content is generated at least in part by AI, but a whopping 96% of developers don’t trust AI generated code.

This means that, at least for now, the 70/30 rule is still intact. But people’s relationship with AI could soon cross that line between aid and dependency.

What Does “Ethical AI Leadership” Even Mean?

A Gallup poll from earlier this year showed that AI use has skyrocketed in every professional industry. Even 25% of those in the healthcare sector use it frequently.

Perhaps most relevant to this piece is that company leaders are the first to lean on AI, and mandate its use within their company. The poll showed that 44% of leaders frequently (at least a few times a week) use AI, while only 23% of employees said the same. This may have a lot to do with leaders finding more reason to use it than employees.

Interested in learning more about using AI properly at work? Sign up for our live virtual workshop, Practical AI Skills for Everyday Work

So in a discussion about using AI ethically, how can company leaders model its proper use? Unfortunately, “ethics” is a subjective word, but Taylor Ventura, who trains others in the use of AI like ChatGPT, had this to say:

“Others see AI as unethical because they believe it replaces craftsmanship or encourages people to shortcut work that should be done manually. On the other hand, many people view AI as an inevitable technological shift and believe the ethical choice is to learn how to use it responsibly so they can become more productive and competitive.”

“From a leadership perspective, ethical AI isn't about whether AI exists; it's about how people choose to use it. Ethical use means being transparent when AI is involved, protecting sensitive information, verifying AI-generated work before acting on it, and ensuring humans remain accountable for important decisions. AI should support human judgment, not replace it.”

“If a leader delegates important decisions entirely to AI without reviewing the output, that's still a leadership failure. Responsibility doesn't disappear simply because technology was involved.”

Where AI Goes from Here

AI’s future is one fraught with uncertainty and trepidation. Teachers are concerned about children becoming unable or even uninterested in learning and actually retaining new knowledge. And while those still learning how to use AI provide humorous moments – like the Canadian politician who read an AI prompt out loud during his speech – it more often causes dismay.

PBS highlighted a United Nations University report that showed data centers used 448 trillion watt-hours of electricity in 2025 alone; more than all but 10 countries on Earth. That kind of consumption is not only unnecessary, it’s unsustainable.

What started three years ago as a fun tool to use sparingly has evolved in a somewhat dangerous way. If we’re not careful, AI use could have much more dire consequences than job replacement. This can be avoided by leaders showing consideration when using it, and by the general population actually caring when AI gets something wrong.