By Chuck Gallagher — Business Ethics Keynote Speaker and Trainer
TL;DR: Chuck Gallagher, AI ethics speaker and author, argues that AI only improves business decisions when leaders are willing to change their minds, and that using AI to justify a choice already made is selective evidence dressed up as analysis.
Picture the owner of a bakery with three locations. His operations manager spends a week using an AI tool to sort through six months of sales data and online reviews. The picture is clear. Weekday lunch is growing fast. The premium cake line is slipping. And customers keep complaining about wait times at the counter.
The owner glances at the summary and shakes his head. He’s been doing this for twenty years, he says, and cakes are the brand. He asks her to run the analysis again, this time focusing on the upside for cakes. The second report, shaped by the new question, supports him. At the next staff meeting he calls the decision data-driven. His manager knows better. She doesn’t say a word.
As an AI ethics speaker and author, I’ve watched some version of that meeting play out in companies of every size. Now there’s data on it. Atlassian’s State of Product 2027 report, built on a Wakefield Research survey of 1,000 product professionals in the United States, Germany, and France, found that 77% say executive gut instinct overrides data-driven product decisions at least sometimes. The same survey found that while 77% of teams formally log customer feedback, 64% say that feedback isn’t consistently used in decisions.
What Is the HiPPO Problem in Business Decision-Making?
Atlassian gives it a name that’s easy to remember: the HiPPO problem, short for the Highest-Paid Person’s Opinion. It describes what happens when the boss’s view crowds out the data and the other voices in the room. The report doesn’t soften the point. It tells executives, in effect, to get out of their own way.
The numbers back that up. Only 4% of respondents said every roadmap decision can be traced back to customer feedback or evidence. Among the obstacles respondents cited, 37% named conflicts with leadership priorities and 43% said they lack a standardized process for using feedback in decisions at all. Teams are collecting the information. Somewhere between the spreadsheet and the decision, it gets lost.
Why Does One Opinion Carry So Much Weight in a Small Business?
Atlassian surveyed companies with 500 to 10,000 employees. In a business that size, a HiPPO has to win arguments against vice presidents and department heads. In an owner-led small business, there often isn’t an argument. The owner’s opinion sets the prices, picks the hires, approves the marketing, and decides how customers get treated. That can be a strength. Owners know things no dashboard captures. But when one person’s view is the only one that counts, a blind spot doesn’t get corrected. It gets multiplied across every part of the business.
Is Using AI to Justify a Decision Already Made an Ethical Problem?
Yes. And AI makes it easier than ever. A leader who has already decided can simply keep asking until the tool agrees. Rephrase the question. Narrow the data. Ask for the strongest case for the option you prefer. These tools are built to be helpful, and they’ll usually give you what you asked for. The result looks like analysis. It’s really a permission slip.
That’s selective evidence, and it isn’t new. It’s the same thing as reading only the reports that agree with you, just faster and more polished. The ethical problem isn’t that the leader trusted his gut. It’s that he presented his gut as something else. When you tell your team a decision was driven by data, and it wasn’t, you’ve told them something untrue. Let me be clear. That’s a credibility problem, and credibility is hard to win back.
Why Do Employees Stay Silent When the Boss Overrides the Data?
Go back to the bakery manager. She had the better information, and she said nothing. Why? Because she’d learned what happens to people who contradict the owner. Nobody had to threaten her. The silence came from the power in the room, not from any rule.
Now, the way I see it, that silence is the real cost of the HiPPO problem. The bad decision is just one decision. A culture where people stop bringing you the truth is a whole string of bad decisions you’ll never see coming. I’ve sat at the head of the table convinced I knew better. Trust me, the people around you notice. And they adjust.
How Should Leaders Document a Decision to Override the Evidence?
None of this means a leader should never trust instinct. Atlassian’s own recommendation to executives acknowledges that instinct is a form of data built on experience the team may not have. The goal is to be ready to change your mind, and to count it as a win when customer insight does exactly that.
The practical fix is simple. When you go against what the evidence recommends, write down three things. First, the evidence, meaning what the data and the feedback actually said. Second, the uncertainty, meaning what the data couldn’t tell you and how confident anyone could reasonably be. Third, your reason for overriding it, stated plainly enough that your team could read it and understand. That record does two jobs. It forces you to be honest with yourself in the moment, and it lets you look back in six months and learn whether your gut was right.
Apply the same scrutiny to AI advice. A recommendation from a machine is still a recommendation. Ask what data it used, what it might have missed, and whether you shaped the answer by the way you asked. AI doesn’t get a free pass any more than the boss does.
As an AI ethics speaker and author, I’ll leave owners with this. AI can put better information on your desk than you’ve ever had. It can’t make you read it with an open mind. That part is a choice. And every choice has a consequence.
Frequently Asked Questions
What is the HiPPO problem?
HiPPO stands for the Highest-Paid Person’s Opinion. Atlassian’s State of Product 2027 report uses the term to describe what happens when a senior leader’s opinion crowds out data and other voices during decision-making. It leads to decisions based on authority rather than evidence.
How often do executives override data with gut instinct?
Often. In Atlassian’s State of Product 2027 survey of 1,000 product professionals, conducted by Wakefield Research, 77% said executive gut instinct overrides data-driven product decisions at least sometimes. The same survey found that 64% say customer feedback isn’t consistently used in decisions, and only 4% say every roadmap decision can be traced to evidence.
Is it unethical to use AI to support a decision you’ve already made?
It becomes an ethical problem when a leader keeps prompting AI until it agrees and then presents the result as objective analysis. Chuck Gallagher, an AI ethics speaker and author, calls this selective evidence: the decision came from opinion, but it’s described as data-driven. Being honest about the real basis for a decision protects a leader’s credibility with the team.
How can a small business owner avoid the HiPPO problem?
Invite disagreement before a decision is made, not after, and make it safe for employees to bring unwelcome data. When overriding a recommendation, document the evidence, the uncertainty, and the reason for the override. Reviewing those records later shows whether instinct or evidence proved more reliable.
Should business owners trust AI recommendations over their own instincts?
Not automatically. AI recommendations deserve the same scrutiny as any other advice, including questions about what data was used, what might be missing, and whether the prompt shaped the answer. Experience and instinct are valuable, but they work best when tested against evidence rather than used to replace it.
Bring This Conversation to Your Organization
If your organization is gathering more data than ever but still deciding the same old way, the gap isn’t technology. It’s leadership. I help executive teams, associations, and conference audiences build decision habits that hold up to scrutiny, including how to weigh AI advice honestly and how to make it safe for people to tell the boss the truth. To bring a keynote or workshop on AI ethics and ethical leadership to your organization, connect with me, Chuck Gallagher, at ChuckGallagher.com.
Research source: Atlassian, The State of Product 2027, based on a survey of 1,000 product professionals conducted by Wakefield Research, August 13–26, 2026.
Five Questions for Reflection
1. Think of a recent decision in your organization. Could you trace it back to evidence, or did it mostly reflect one person’s opinion?
2. When was the last time someone on your team told you something you didn’t want to hear, and how did you respond?
3. Have you ever reworded a question to an AI tool, or to a colleague, until you got the answer you wanted?
4. What would change if every decision that overrode the evidence had to be explained in writing?
5. Where is the line between trusting experienced judgment and ignoring inconvenient facts?
