Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

44 leaders on whether AI belongs on the team

Where accountability, judgment, and human oversight still matter.

[Photo: Getty Images]

Profile photo of Fast Company Executive Board

Advertisements for artificial intelligence are pushing social connection, treating agentic AI like a helpful neighbor or a cool friend that “gets” you. In business life, human resource terms once reserved for people—employee, colleague, team member—are now being applied to AI. Whether these monikers stick remains to be seen.

We asked our Fast Company Executive Board members whether they consider AI a team member and why. A quick glance at the responses shows there’s no clear winner between yes, no, and not entirely. Nearly everyone, however, cautioned that final accountability should remain with humans. Here, 44 members who responded share whether they think of AI as a team member and how it fits within their organization.

1. WHERE IDEAS GO BEFORE THE TEAM SEES THEM

Yes, Claude is where my ideas go before my team sees them. I use it to draft, argue with myself, and catch the lazy thinking before it reaches a client or a deadline. It works as a sharper mirror for my own thinking. — Sharon Lee Thony, SLT Consulting

2. AN ASSISTANT ONLY

No, AI is a powerful tool and force multiplier, but technology doesn’t solve problems—people do. I view AI as an assistant that speeds up execution and handles heavy lifting, but it lacks human empathy, intuition, and authentic storytelling. If you rely on AI to be the creative engine or strategy, it quickly turns into “amateur night”. Real business impact requires human direction, critical thinking, and heart. Use AI to augment your team, not replace the human spark that builds real connections. — Kristi Melani, Reputation

3. STILL JUST A MACHINE

No. It’s a tool. Even if something runs autonomously, it’s still a machine. I know there’s a growing tendency to talk about AI agents as employees or digital coworkers, but I think that language gives the technology too much credit and can blur accountability. AI can execute tasks, analyze information, and increasingly make decisions within defined parameters, but a human still must decide what “good” looks like, where the guardrails are, and who is responsible when something goes wrong. I use AI constantly, but I don’t manage it like a person. I use it like an increasingly capable piece of technology. — Andrea Lechner-Becker, GNW Consulting

4. VALUABLE THOUGHT PARTNER

No, I don’t consider AI a team member, but I do see it as a valuable thought partner. I think of it as a personal mental organizer—most effective when used consistently enough to understand the questions you ask, the patterns you see, and the context behind your work. Its value is in helping synthesize information, surfacing connections, and identifying opportunities faster. Human judgment, lived experience, intuition, and context are what the team brings to the table. The AI counterpart is more like a highly responsive consultant that helps us think and organize more expansively and more rapidly. — Melissa Gonzalez, MG2

5. BETTER TOOL TO WORK SMARTER

A startup may eventually have one employee and a hundred AI agents running different tasks, but that does not mean it has a 101-person team. AI can dramatically extend what employees are capable of doing. I see its greatest value in making people more productive, informed, and effective. Better tools allow everyone to work smarter. Calling AI an employee also blurs an important professional distinction: AI does not carry accountability for its decisions or consequences. People do. The technology may perform more of the work, but responsibility still must sit with the humans using it. — Goran Paun, ArtVersion

6. PART OF A “TOOLS-TO-TEAMMATES” MINDSET

AI should be considered a teammate, not simply a tool. To become AI-first, organizations must foster a “tools-to-teammates” mindset, where AI is embedded into how work gets done and works alongside people to improve outcomes. That doesn’t mean replacing human expertise. AI can take on execution, provide context, and accelerate decisions, while people remain responsible for judgment, creativity, and accountability. The opportunity is to intentionally design work so humans and AI complement one another, allowing employees to focus more of their time on the uniquely human contributions that create value. — Guillaume Lardeux, Genesys

7. THE ADVISOR WITHOUT AUTHORITY

I consider AI an advisor, not a team member. It can surface patterns, generate options, and accelerate analysis, but it lacks lived context, institutional accountability, and authority to make consequential judgments. Its outputs should be tested against reliable evidence, challenged for bias or error, and validated by a qualified human who remains responsible for the decision. — Britton Bloch, Navy Federal Credit Union

8. EXCEPTIONAL TOOLS, NOT TEAMMATES

No, and I think framing it that way is a mistake. AI tools are exceptional tools, but they’re not teammates. Calling them teammates muddies who’s actually accountable for the work and the outcome. We use AI aggressively across the business, and it’s made us dramatically more capable, but the people are still the people. — Zander Cook, Lease End

9. THE INTERN-LEVEL RESEARCHER

We think of AI as a terrific intern and research assistant. It’s great at synthesizing information and saving time. Like a talented intern, we are very specific about its capabilities and limits and don’t ask too much of it. It’s not a senior team member and it’s not capable of building trust, which is the core value proposition we offer. — David Meadvin, One Strategy Group

10. TOOL TO AMPLIFY TEAM MEMBERS

No, AI can’t be on the team for the simple reason that it isn’t a person. But we do use AI to amplify every single team member. Our people have access to the latest models and the budget to use them heavily. And we run our company on a knowledge base powered by GitHub and Claude Code. Everyone updates it every day with their work as code, documents, and visuals, and it’s all available across the company. This means anyone on the team can ask AI questions about what’s happening in the organization and get up to speed on projects without interrupting their colleagues. When we do meet, we can focus on strategy instead of logistical updates. — Michael Quoc, Product.ai

11. A “NAMED” TEAM MEMBER

We began referring to AI by “people names” as a joke, but it’s become more of a reality as AI has evolved into a true team member. Now we use those names to tag it in workflows, assign it tasks in project management tools, and more. — Jen Jones, Siteimprove

12. THE ANALYST WHO NEVER TAKES THE HEAT

Yes, with a caveat. A team member carries accountability. AI carries none, and my name still goes on the recommendation. So, I treat it like the sharpest analyst I’ve ever worked with, one who never sits in the room and never takes the heat. That shapes how I use it. I bring it my thinking and tell it to find the holes rather than asking it for answers. Half the time it finds the crack I talked myself past. The other half is memory. I’ve built a second brain that tracks every project I’m carrying and surfaces intersections I would have walked past. It hasn’t replaced a thing I do. It has raised the floor on everything I ship. — Trent Cotton, ICIMS

13. POWERFUL COGNITIVE TOOL

What constitutes being a team member? Is it simply the ability to be competent at a task, or is it something more? I find this more a philosophical question rather than a technological one. One of my favorite authors, Iain M. Banks, explores this concept of sentience in his Culture series. Sentient machines must cross a threshold where, instead of asking if they can produce intelligent-looking outputs, become an entity with their own interests, autonomy, and independent agency. Until AIs cross this threshold, I choose to consider them as a powerful cognitive tool versus granting them personhood as a team member. — Anurag Patnaik, Autodesk

14. DIGITAL DELEGATE

I view AI as a digital delegate rather than a team member or employee. Treating AI like an employee creates risky autonomy and unmanaged costs because software has no skin in the game when things go wrong. It lacks a moral compass or any understanding of risk. You can delegate low-risk tasks to AI, but establishing human oversight for execution and escalation is key when it comes to high-stakes issues. A human team member has legal liability, intent, and professional judgment, and is who should be ultimately responsible for business outcomes. — Faisal Masud, HP

15. AUGMENTED INTELLIGENCE COMPANION

Absolutely. I see AI not as a search engine or mere generative guesswork, but as an active, agentic team member. We built our AI-driven platform using deterministic neural “cells” that perform complex tasks exactly as a highly trained human colleague would from calculating income to processing loans. It doesn’t just guess. It analyzes facts and automatically loops in a human counterpart when it encounters something it doesn’t know. In this way, AI isn’t replacing the team. It is an augmented intelligence companion that works alongside us 24/7 to eliminate bias and empower our originators. — Pavan Agarwal, Celligence

16. TEAM MEMBER WITH CAVEATS

In many ways, AI is a team member. It can take on work, surface insights a human might miss, and meaningfully extend what a team is capable of. But just like any team member, you need to understand its capabilities and gaps before handing off real work. Not every AI use case requires the same level of human oversight. Some may need only a light review, while others require a much more robust check before anything moves forward. The ones getting it right define those guardrails early, by use case, revisit them consistently, and adjust when something is not working as expected. The technology will keep evolving, but the oversight must keep up. — Courtney Dutter, ICIMS

17. CO-WORKER, NOT TEAM MEMBER

Is AI a team member? I get the point, but I think regarding it as a team member is a slap in the face of the humans on the team. Team members share a purpose, devote passion, and sacrifice. And AI—so far—is capable of none of those things. Co-worker, yes. Colleague, yes. Thought partner, sure. But not a team member. — Bob Sprague, Yes&

18. IT’S LEVERAGE

AI is not a team member. It is leverage, and the difference is accountability. A team member owns a recommendation and answers for it when it lands badly. No model does that. The harder problem shows up later. Our young professionals learned judgment by writing the bad first draft and getting it marked up. Take that away and you get a bench that can polish fluent copy but cannot tell you whether it is right. Critical thinking must keep evolving. I drove out West, where there was no signal, and it was a good thing we still knew how to read a paper map. Use AI where the thinking is already done. Protect the part where it isn’t. — Kathleen Lucente, Red Fan Communications

19. SEAT ON THE TEAM, WITH LIMITS

AI has earned a seat on the team, but one that doesn’t get credit and can become a liability. AI can move faster and operate at greater scale than humans, but its value depends entirely on the quality, security, and governance of the data behind it. Because AI cannot own judgment or accountability, every agent needs a clear purpose, guardrails, measurable outcomes, and a human sponsor. The companies that win with AI will be those that build trust by design, starting with trusted data. — Shiva Pillay, Veeam Software

20. LIKE A TOOL FOR TEAMS

No, I treat it like a tool a team member uses, not a team member itself. The difference matters because tools don’t own outcomes, people do. The second you start thinking of AI as a colleague, accountability gets fuzzy. On our projects, a developer using AI to move faster is still the one responsible for what ships. That framing keeps quality standards where they belong and stops people from blaming the model when something breaks. — Daniel Haiem, AppMakers USA

21. THE THOUGHT PARTNER THAT DOESN’T HAVE YOUR BACK

Not yet. The level of trust I have in a team member isn’t there yet with AI. I use it as a quick-access thought partner, faster than pulling someone else off their own work, to challenge my thinking, offer a different perspective, or point out a blind spot. It’s valuable when I need to get unstuck, but does it have my back the way a team member does? Not yet. — Geri Johnson, Next PR

22. THE FORCE MULTIPLIER

I think of AI as a force multiplier and an occasional sounding board rather than a teammate—it doesn’t share accountability the way a person does. We use AI tools to move faster on diligence and pattern recognition, but every judgment call still runs through a human who owns the outcome. The tools constantly become that much smarter, but, ultimately, the judgment and accountability stay with the human being. — Elliott Broidy, Broidy Capital Holdings

23. THE RESEARCH ASSISTANT

AI runs my structured-data audits, cross-checks facts across platforms, and drafts first-pass content I edit before publishing. This year it helped me audit 18 profiles with my name on them and flag every place the facts didn’t match. It doesn’t own outcomes or make decisions without me, so I don’t count it as a team member. I count it as a research assistant that works in minutes instead of days. — Hasan Saleem, DSS MEDIA

24. MAJOR MULTIPLIER

Not yet. A human brings historical experience, business context, and continuous learning accrued over time in their role. An AI agent only has the knowledge you give it. Our research with Wakefield found 77% of leaders say less than 20% of their data is contextualized enough for agents to use. Until that gap closes, agents can’t be trusted to act autonomously. But when you give agents context, governed data, and guardrails and implement learning mechanisms, they can be a major multiplier for the team. — Sumeet Arora, Teradata

25. LEVERAGE FOR OUR TEAM

I do not treat AI as a team member because it cannot own an outcome, read a client relationship, or be accountable for a decision. I treat it as leverage for our team: It can accelerate research, first-pass outlines, and administrative work, while people remain responsible for strategy, creative judgment, and final quality. Tools can increase capacity, but leadership and accountability must stay human. — Tavares Beverly, Beverly Boy Productions

26. THE HIGH-POTENTIAL INTERN

I view AI as a high-potential intern. We utilize it to handle mechanical work so human teams can focus on innovation, creativity, and job satisfaction. For example, we task AI with automating tedious code and backend maintenance, eliminating hidden digital downtime and freeing engineers to focus on high-impact innovation. The work gets done, the team doesn’t get burnt out, and the human focus stays on meaningful changes that improve the customer experience. — Benjamin Spiegel, POLYWOOD

27. TRUE CONTRIBUTING TEAM MEMBER

If you want AI to be a true contributor rather than a souped-up search engine, you have to treat it as a member of the team. Because AI retains context, it can build upon prior context to offer increasingly sophisticated answers. And the more team-specific data you provide it right from the start—strategic plan, communications, etc.—the better it will perform. Just ensure it’s a secure platform, your own Copilot, for example, rather than an open one. — Shani Harmon, Stop Meeting Like This

28. USEFUL TOOL FOR CURATION

AI is not a team member, just as your keyboard or trackpad isn’t. Team members have feelings and unique experiences that drive great creative work. AI doesn’t have a life of experiences. Rather, it has an extraordinary ability to remix existing content. It’s a genuinely useful tool for curation, not creation, and we do our people a disservice by blurring that line. When creating with my team, I’m not looking for any algorithmic or statistical output. I’m looking for what happens when people with different life experiences collaborate. AI can accelerate that process, but calling it a team member risks outsourcing the human heart of the idea. — Don Jeter, Torq

29. POWERFUL CAPABILITY MULTIPLIER

I see AI as a powerful capability multiplier, not a team member. It can help people analyze, create, and move faster, but it does not replace human judgment, accountability, or an understanding of the wider business context. For leaders, the real question is not whether AI belongs on the team. It is whether the team has the confidence and capability to use it well. As AI changes work, organizations need to accelerate speed-to-skill, rather than assuming people will simply figure it out alone. — Nick Gonios, Epignosis Learning Technologies

30. DIGITAL EMPLOYEE, WITH EXCEPTIONS

Yes, but with exceptions. I think of AI as a digital employee, not a piece of software you switch on. Like any new hire, it needs a clear job, the right context, boundaries on what it can decide, and a way to measure whether it’s doing good work. The difference is a person absorbs context from meetings, side conversations, and watching how the team operates. AI can’t . . . yet. If your process lives in people’s heads, especially the exceptions and judgment calls, the agent will struggle. When an AI deployment stalls it usually isn’t, “Is the model good enough?” It’s “Have we given it a clear job description and taught it how this company works?” — Joe Addona, Vida Global

31. HIGHLY CAPABLE INSTRUMENT

I consider it not a team member, but a highly capable instrument. In neurosurgery, AI already supports imaging interpretation, surgical planning, and outcome prediction, but it doesn’t carry accountability. A resident, a nurse, or an attending clinician owns a decision. AI informs it. The moment we treat it as a colleague, we blur the line of responsibility, and in medicine that line has to stay sharp. — Steven Giannotta, Giannotta Neurosurgery

32. POWERFUL COLLABORATOR

I consider AI a powerful collaborator, but not a team member in the human sense. It can accelerate research, analysis, and execution, but accountability, judgment, and purpose must remain human. The leadership challenge is not simply learning how to use AI but knowing what thinking we should never outsource to it. — Stoyana Natseva, Happy Life Academy

33. HIGHLY CAPABLE TEAM MEMBER

I do. I think of AI as a highly capable team member that can help with research, notetaking, drafting email and other communication, and organizing information. Because your name and reputation are still behind any deliverables that you may leverage AI to produce, AI still needs professional oversight for approvals. But when used well, it can save significant time and allow people to focus more energy on judgment, building relationships, and financial planning. — Ryan Walker, Meristead Wealth

34. AN EFFICIENCY TOOL

No, I consider AI a tool that helps our team members to be more efficient. We have eliminated a few administrative positions because the work could be completed faster and more accurately by an AI tool. However, this isn’t any different from other technology advances that reshaped the workplace over the years. — Sheri Jacobs, Avenue M Group

35. POWERFUL TOOL FOR IMPROVED WORKFLOW

No, I don’t consider AI a team member. It’s a powerful tool, but it still can’t be trusted with human decisions that require judgment, context, accountability, and an understanding of consequences. I’m comfortable using AI to support research, speed up workflows, and help teams process information, but the final call still needs to come from a person. Until AI can reliably handle nuance and responsibility at a human level, I see it as an assistant to the team, not a member of it. — Michael Gargiulo, VPN.com

36. THE ALWAYS-ON COLLEAGUE WITH NEEDED GUARDRAILS

AI can be an effective team member with the right guardrails in place. A general-purpose tool may require excessive oversight, creating more frustration than value, with little return on investment. Just like a human team member, AI has to be trustworthy. Domain-specific AI that is purpose built for your team’s actual use cases can be an incredible “always-on” colleague—proactively surfacing the insights that matter and handling repeatable workflows that you can trust and act on immediately. — Kiva Kolstein, AlphaSense

37. TEAMMATES WHO COLLABORATE

Yes, I consider AI a team member 100%. And in the office, we (the humans) talk about them as if they are teammates. Each of our AI tools comes with unique qualities. Claude (Fable 5) is our “get-stuff-done” workhorse. Perplexity is the office researcher (the trusted aggregator of data). ChatGPT, when em dashes and other AI tells must be accounted for, is our editor. Granola is our note taker. And, even better, they all communicate with each other, collaborate, check each other’s work, and provide feedback. I honestly can’t even imagine working without our AI team. It’s one of the most critical ways we’ve been able to scale. — Tia Katz, Hu-X

38. TEAM MEMBER YOU HAVE TO DOUBLE CHECK

Yes, I treat AI like a team member, but one I have to double check constantly. It’s fast and it’s smart, but it will lie to you with total confidence if you let it. I never publish or send anything AI touches without verifying it myself first. — Kevin C. Roy, GreenBanana SEO

39. POWERFUL TEAMMATE, UP TO A POINT

I think AI can be a very powerful teammate, but it shouldn’t replace the people closest to the experience. In a restaurant, so much happens in real time—a guest deciding what to order, a server reading the table, a manager trying to solve a problem. AI can bring intelligence into those moments, giving people insights they can act on while it still matters. To me, that’s where the real opportunity is. — Rhonda Levene, Ziosk

40. A RESEARCH ASSISTANT ROLE

Absolutely I consider AI as a team member, but more in the role of a research assistant than an executive. When we’re working with a client undergoing a significant change in how they operate from a brand standpoint, we often look for analogs—companies that have gone through something similar so we can learn from their successes and mistakes. Previously, finding truly relevant examples could take days or weeks. Now a simple prompt (or three) can surface a dozen potential analogs almost instantly. The catch is some of those examples may be AI hallucinations or irrelevant. But if five out of 12 are genuinely useful, that can still save an enormous amount of time. — Bruno Benedini, Taillight

41. CAPABLE TEAMMATE TO SYNTHESIZE INFORMATION

At my company, AI continuously analyzes building, occupancy, environmental, and operational data to identify patterns, surface opportunities, and increasingly help optimize building performance. It can do work that would be difficult for a person to do continuously across thousands of building data points, while building operators remain in control of the decisions and outcomes. Think of it as a highly capable teammate: AI provides continuous analysis, recommendations, and automation. People provide judgment, context, and accountability. — Jennifer Nuckles, R-Zero

42. TOOL TO AUGMENT WORKER ABILITIES

To me, AI is a tool that can augment worker abilities. Artificial intelligence helps workers to make smart decisions by handling data location and compilation and making recommendations. But I stop short of trusting AI to execute without human review. AI can also help writers draft effective content by concepting article ideas, organizing key points, and creating a rough draft. All content should be reviewed for errors and rewritten into the appropriate voice before being shared with others, however. This describes a type of augmented intelligence for workers—a tool—rather than the capabilities of a team member. — Christina Robbins, Digitech Systems, LLC

43. A SOPHISTICATED MIRROR

Framing AI as a team member is dangerous. A team member owns outcomes and can be held accountable when things go wrong. AI can’t do either, so the minute you call it a teammate, you’ve given people a convenient place to park accountability. I treat AI as a capability that makes my team faster, with a named human accountable for every output it touches. I’d also challenge the “thinking partner” label. Most people use AI to validate what they already believe, which turns the model into a sophisticated mirror rather than something that stress-tests their judgment. — Derrick Pledger, Maximus

44. ON MY PAYROLL, NOT IN MY FOXHOLE

AI is on my payroll, not in my foxhole. We built an empathy engine that hundreds of thousands of young people told things they would not say to a parent or a teacher. So, I know exactly how much a machine can hold and where it stops. What I keep coming back to is that it has no skin in the outcome. It has never once called me at 11 p.m. to say the thing that every human in the meeting was too polite to say. Treat it as a colleague, and sooner or later you will ask it for judgment. I have watched it produce something confident and hollow almost every time. — Natanya Wachtel, New Solutions Network