If your company is like most, you probably use several AI tools. The finance department may be using artificial intelligence (AI) to streamline budgets. Sales might be trying out an AI agent to generate leads. And HR might be using AI to source candidates. It’s all exciting and good — until projects overlap and no one knows which are the “right” tools to use.
The solution? An AI committee. This governing board can oversee your company’s AI usage, develop a strategic vision, reduce risks, and guarantee that the technology is used in an ethical way. In short, it’s a clearinghouse for all things AI.
Here’s what you need to know about AI committees, why you might want one, and why you might not.
What is an AI committee?
An AI committee is a governing body that guides the vision, strategy, and implementation of AI in your company. Many companies already have such a committee but they call it an AI council, center of excellence, or advisory board. An AI committee also makes sure a company deploys AI in a responsible and ethical way.
What does an AI committee do?
The mission of an AI committee can vary from company to company. But overall, a committee inventories how its company uses AI and makes sure that teams aren’t duplicating each other’s efforts. It asks the right questions to make sure AI tools solve real business problems and further the company’s objectives. And it stays abreast of AI regulations to keep the company in compliance.
Who should be on an AI committee and how should it be structured?
No matter what you call your AI committee, it needs to include people with a wide range of expertise, such as those from the IT, legal, finance, ethics, data analytics, HR, business, and security and compliance departments.
“It’s really important to have HR on the committee,” said Ashleigh McCord, Salesforce senior vice president, strategy and enablement, “because you want to think about the workforce implications of AI — how to manage the people aspects of this, and how to reskill your employees.”
Business leaders need to be involved to align AI tools with the company’s objectives. And C-suite members should be included too — or, at the very least, you should secure their buy-in. “At Salesforce, we have multiple C-suite players at the table,” McCord said, “including the chief digital officer, chief financial officer, and chief product officer.”
AI committees can be structured in several ways. Some companies have one entity that oversees everything. Others, like Salesforce, have several. Salesforce’s Customer Zero SteerCo, which McCord leads, handles the big-picture strategic AI vision for the company, while its AI Council makes sure all tools meet legal, compliance, and security requirements. As for the final buying decisions? “It should be IT making that decision,” McCord said, “along with the business leaders.”
So, do you need an AI committee? Consider the pros and cons below.
5 reasons to have an AI committee
AI committees can help your company in many ways. Here’s a list of the top reasons why you should consider creating one.
1. An AI committee can survey how AI is being used in the company
You may not have the full picture of how — and where — your employees are using AI. An AI committee can catalog this by asking: Which teams and departments are deploying AI? What are they using?
With this information, the committee can create an inventory of your AI tools, document use cases, assign ownership for each tool, and maintain the roadmap. But it also needs to assess any unauthorized usage in the company.
Make no mistake, your employees are using AI whether you allow it or not. According to 2024 research from Cyberhaven, 74% of ChatGPT use at work is through noncorporate accounts; use of Google Gemini is higher, at 94%. All that unauthorized AI access to your corporate data, including legal documents, HR records, and other sensitive information, puts your company at risk.
Once the committee has a complete picture of how your employees are using AI, it can decide what’s working and what’s not.
2. It can align AI tools with business goals and create usage guidelines
One of the biggest mistakes companies make with AI is that they rush to adopt a tool before they identify the problem they want it to solve. An AI committee can hit the brakes and ask, “How does this tool further our company’s goals?”
But it can go even further and proactively seek AI solutions to your company’s biggest pain points. The committee might look at gaps or challenges that exist within the company, and suggest AI tools to solve them. For example, if a resort wanted to improve guest satisfaction scores by 15%, the committee could recommend an agentic AI solution, such as Agentforce, the agentic AI layer of the Salesforce platform, to personalize interactions with its guests.
An AI committee can also create guidelines of when, where, and how AI should be used in the company. It can do this by developing principles for responsible AI to clarify your stance on fairness, transparency, and accountability.
Need advice on how to do all of this? Check out the World Economic Forum’s comprehensive guide to creating an AI governing body.
8 questions to help you start an Agentforce use case
You need a solid approach to make sure AI agents will work for your team and customers. This game plan maps out how Agentforce can improve your workflows, eliminate friction, and achieve real results.
3. An AI committee can break down silos and create a culture of collaboration
When different teams deploy different AI tools, companies end up supporting a variety of infrastructures, tools, large language models (LLMs), and approaches to scaling. According to McKinsey, too many platforms is one of the top obstacles to implementing AI at scale.
This is where an AI committee comes in handy. It can look at what each team is doing, identify duplications, and then suggest a more efficient solution. “Our SteerCo focuses on this streamlined approach and breaking down some of those silos,” said McCord. “We might say to several different teams, ‘Did you know we have this data or capability? What if we brought those two things together?’”
Not only does this foster collaboration and reduce redundancy, it also saves the company money.
4. It can help reduce risk and ensure compliance
If you have an AI committee for no other reason, consider this one: Companies need a central governing body to make sure they’re in compliance with local regulations and to protect themselves against risk.
Once your AI committee has made a complete inventory of the company’s AI tools — and decides which ones it wants to use — it can brainstorm potential risks. What are the security vulnerabilities of a particular tool? Is it trustworthy? Could there be unintended consequences of the tool’s actions? The advantage of having a broad-based committee is that legal might identify one set of concerns, while IT comes up with others. The committee can then assess how likely these risks are to occur and prioritize them according to their potential harm.
The committee can also help the company stay in compliance with regulations. Right now, regulations for AI are a patchwork of emerging guidelines and policies that vary by region and industry. What’s true in Europe or Africa, for example, may be different in North America. And because governments and regulatory bodies are racing to stay ahead of technological advances, it can be hard to predict what regulations will look like even six months from now.
By having one governing body that stays abreast of the regulatory landscape, you’ll benefit the whole company.
5. An AI committee can measure how well your tools are working
A central governing body can also measure outcomes of the projects that have been approved. Going back to the example of the resort: If the goal was to increase customer satisfaction by 15%, did the AI agent make that happen? What do the scores look like now?
“It’s really important,” said McCord, “for committees to be able to show results, measure usage, and see how projects are performing.” Each tool needs to prove its business value, as well as its technological proficiency. When a tool or platform is working, the AI committee can continue to give it the green light. When it’s not, the committee can refine its strategy — or ditch the tool altogether.
Is your company ready for an AI committee?
Before you get started, consider if your infrastructure and even your company culture is ready to take this step.
How much prep work has your company done?
AI can only perform as well as the the data and structure it’s built on, which is why Joe Peppard, a professor at the UCD Michael Smurfit Graduate Business School in Dublin, argued in a recent Wall Street Journal article that most companies aren’t ready for an AI committee, let alone an AI strategy.
“Even if a well-crafted AI strategy were to somehow magically land in a CEO’s inbox, the organization likely won’t be able to implement it,” Peppard wrote. “Most just haven’t done the foundational work.”
According to Peppard, companies need to clean up their data, making sure it’s free of errors, incorrect formats, duplicates, and mislabelings. They should also modernize their tech infrastructure, increase digital literacy, and put guardrails in place. Only after that’s done, he wrote, is a company ready for the next step.
How fast does your company move?
Quick, what’s the one phrase that’s most likely to kill creativity or innovation? If you said, “Let’s have the committee review it,” you’re probably right.
AI committees can occasionally stop innovative solutions from being deployed. If someone from sales comes up with a great idea, for example, the AI committee could turn it down because that department already has too many tools. “Some people could look at that and say that it stifles innovation, but that’s not the objective of an AI committee,” McCord said. “The objective is to try and get the right tools adopted in the right flow of work.”
Because an AI committee serves as a funnel at the top of the company, it can also create bottlenecks. If the committee approves or denies projects quickly, that’s great. If not, it can throw a wrench into a fast-moving tech landscape. In other words, if your company wants to move fast with AI, a committee may not be the best choice.
But even if an AI committee has its downsides, its benefits are greater. “Obviously, it takes time when you go through a committee, so that could be seen as a negative,” said McCord, “but it also ensures that we are implementing AI in a trusted, secure, and appropriate way.”
D-I-Why? Deploy AI agents faster with Agentforce
Building and deploying autonomous AI agents takes time. Agentforce, the agentic layer of the Salesforce platform, can reduce time to market by 16x compared to DIY approaches — with 70% greater accuracy, according to a new Valoir report.