AI Ethics Explained: Preparing for the Human-AI Future

**Alexis Hueringa** (0:00)
Welcome to Gartner ThinkCast. I'm Alexis Hueringa. As AI becomes embedded in more business decisions, workflows, and employee experiences, organizations are facing a new challenge. It's no longer just about how to adopt AI, it's about how to do so responsibly. From concerns about over-reliance on AI, to questions around privacy, trust, and human rights, leaders are finding that ethical considerations are becoming business considerations. So how can CIOs and technology stay ahead of these challenges without getting caught in a cycle of reacting to the latest AI controversy? To help unpack all of this, we're joined today by Gartner Senior Principal Analyst, Shanna Grafeld. Shanna, welcome to ThinkCast. Thank you so much, Alexis, for having me on.
So to get right into it, you spent a lot of time researching AI adoption, organizational behavior, and AI ethics. What's driving the increased focus on ethics as organizations accelerate their AI investments?
Sure. So right now, a lot of organizations have been experimenting with AI for several years. And it's starting to move from the edge of the organization, maybe in innovation labs, more into core operations. And as any technology moves from the edge into the core of your organization, it's going to bring with it novel risks. And with AI, there are a lot of ethical concerns that often catch organizations off guard. So things that are ethical considerations and concerns are really becoming strategic business level concerns. And the expectations are increasing. Employees, customers, internal and external stakeholders, they've all got high expectations and are expecting AI to be used in a way that is ethical, that makes people more capable and better, and isn't seen as taking away from human capabilities in the workforce. So leaders are being asked a whole suite of questions around ethical AI that they just weren't being asked a few years ago when they started their AI rollout and it's catching a lot of people off guard and forcing them to be reactive instead of proactive.
Yeah, as you mentioned, organizations are tending to approach this reactively rather than proactively. Can you explain a little bit more about that? What does that look like? Sure. So the first thing to remember is that this is not new. We've seen the same challenge over and over again throughout all of human history. We invent and scale a technology before we have any guardrails or governance around it. And that creates this very dangerous lag time between when we start using a technology and when we actually have governance in place for it. So cars, for example, cars began to be sold commercially in the United States or in the 1890s. But the first law banning drunk driving, for example, wasn't passed until 1906 And we didn't have a national blood alcohol content standard until 2000 And in that time, there were thousands of injuries and deaths caused by drunk drivers. So when it came to cars, we had a 100 year lag time between when cars were available for sale and when governance was standardized and fully enforced.
So right now with AI, we are just at the very beginning of what would be this kind of dangerous lag time. And the problem is we're historically really bad at determining what guardrails are needed for new technology until bad things happen. So for example, we're just now getting caught up with ethical considerations for social media. And that technology is over 20 years old now. So compared with the car example, we're hopefully getting better and faster at this.
But we're constantly being kind of caught off guard because we don't know what ethical challenges we should be looking out for with AI until they're happening to us. Yeah, that's a little concerning thinking about how early on we are still and how those gaps happen. Gartner has also been researching the broader human-AI relationship. How does that perspective change the way leaders should think about AI adoption?
Sure. So there's a lot of research happening right now around what it means to be a human in the age of AI is shifting. And AI is becoming a participant in business processes in many organizations, not just a tool. So we're having to clarify what and explore as we go, right? So we're clarifying it, but we're also figuring it out as we go what this relationship should be. So as a participant in business processes, it can make the lines a little bit blurry between participant and tool. And this is that frontier that organizations are still figuring out. Because we know that, okay, let's say you try to manage AI like a human employee. You put it in the org chart. Well, you're going to trigger mistrust. So you probably don't want to do that. But AI in many organizations is still becoming more than a tool. So you also don't want to necessarily manage it like all of your other software solutions. And employees are looking to leaders to have the answers and be able to create this relationship.