AI Has a Human Foundation: Your ERGs Have Been Integrating Humans for Forty Years

For everyone who works in a company right now.

AI is built on a human foundation. ERGs are human groups. ERGs have been integrating new and diverse talent into organizations for four decades. They know how to do it. The companies that integrate AI well will be the ones that put their ERGs in charge of doing it.

Most companies are not failing at AI because the tools are bad.

They are failing because they are trying to install AI the same way they would install new software, and AI is not behaving like new software. It is behaving like a new group of people walking into the building. People who do not yet know where they fit, what they are allowed to do, or who has their back.

If that sounds familiar, it should. Companies have been through this before. Many times. And the people in your company who already know how to handle it are sitting in a part of the organization that almost nobody is paying enough attention to.

What is actually going wrong with AI

Walk through any company right now and you will see the same pattern.

Some teams are using AI all day. Some teams are using it in secret because they are not sure if they are allowed to. Some teams are refusing to touch it. Some leaders are pushing it hard. Other leaders are quietly hoping it goes away. The training did not really land. The policies are confusing. People are nervous about saying the wrong thing about it in a meeting.

Most executives look at all of this and say the same thing. We need better training. We need clearer policy. We need a better rollout plan.

Those things help. But they do not fix the real issue.

The real issue is that AI is being treated like a tool, when it is showing up like a new member of the team. And new members of a team need something different than tools need. They need to be welcomed. They need to know where they fit. They need someone to vouch for them. They need time to earn trust.

The two questions that decide everything

Anyone who has ever been the new person in a workplace knows what these two questions feel like, even if they have never said them out loud.

Do I belong here? Am I really seen here?

The veteran who left the military and walked into a corporate office for the first time asked these two questions. The new mom coming back from leave asked them. The first Black engineer on the team asked them. The first openly gay manager asked them. The employee in a wheelchair who needed the meeting room rearranged asked them. The first person from a non traditional school in a room full of Ivy League grads asked them.

They are simple questions. They are also the questions that decide whether someone stays, contributes, and grows, or whether they shrink, leave, or quietly check out.

AI is asking those same two questions inside your company right now.

Does AI belong in this meeting. Does AI belong in this customer email. Does AI belong in this hiring decision. Who decides. Who is allowed to use it. Who is going to get in trouble if they do. Who is going to get in trouble if they do not.

How is AI seen in this company. As a threat to my job. As a shortcut. As a cheat. As a partner. As something the company actually trusts. As something the company is just experimenting with so it does not look behind.

Until those questions get clear answers, everything else stalls. The training does not stick. The pilot programs do not spread. People stay quiet about how they are actually using it. And the company keeps spending money on something that is not landing.

What forty years of ERGs have actually taught us

To understand why ERGs are the right answer to the AI question, it helps to understand why ERGs exist in the first place, and why some of them have lasted four decades while others quietly disappeared.

ERGs were not invented by HR. They were built by employees. The first ones came together in the 1970s, when Black employees inside companies like Xerox started meeting on their own time, without a budget and without permission, because the company they worked for was not built for them. They needed a way to be heard. They needed a way to support each other. They needed a way to translate what they were experiencing into something the company could actually act on.

That is the original DNA of an ERG. It was never an event committee. It was a group of employees solving a problem the company did not yet know it had.

What has made ERGs sustainable over four decades is not warm feelings. It is value. The ERGs that have survived and grown are the ones that produced something the company could not get anywhere else.

Recruiting pipelines into talent pools the company could not reach on its own. Retention insight from people who had been considering leaving. Cultural translation that helped the company avoid expensive mistakes. Crisis response when news broke and the company needed to know how its own people were feeling. Customer intelligence from employees who shared identity with the customers the company was trying to serve.

That is the work that earned the ERG its seat. Not the celebrations. The contributions.

The ERGs that lost ground over the years are the ones that stayed only social, stayed only symbolic, or stayed only at the edges of the business. They were the first to lose budget when the company tightened up. They were the first to be questioned when the political environment shifted. They were the first to be asked to justify their existence.

The pattern is clear. ERGs survive when they are useful. They thrive when they are indispensable.

The ERGs still standing today are the ones that figured out how to be load bearing. They moved from affinity to infrastructure. They sat next to recruiting, learning, communications, and product. They became the place the company turned to when it needed to integrate someone, reach someone, or understand someone the rest of the system did not know how to handle.

That is the muscle. And it is the exact muscle the AI moment is asking for.

A note on where I am writing this from

I have spent more than a decade building veteran integration programs inside civilian organizations. Programs that took people whose previous workplace was structured, hierarchical, mission driven, and full of acronyms, and helped them land in workplaces that were none of those things.

That work taught me something specific about ERGs that does not get said often enough.

Veteran ERGs are some of the most undersubscribed groups in the company. Not because veterans are not there. They are. But because veterans often do not self identify, and because the perceived value of joining the group is low if the group has not built something that matters to the business.

The veteran ERGs that have grown are the ones that became useful. They built mentorship programs that retained people. They built referral pipelines that filled hard roles. They built translation work that helped managers understand how to lead people whose first leadership experiences were in a very different system. They became something the company would notice if it disappeared.

The veteran ERGs that have shrunk are the ones that stayed at the edges. They held a Veterans Day breakfast every November and that was most of what people knew about them.

This is not a story unique to veterans. It is the story of every ERG. The ones that survive are the ones that solve a real problem the company has. The ones that fade are the ones that do not.

Right now, every company has a real problem with AI. The ERGs that step into that problem will become the most valuable groups in the building. The ones that do not will keep fighting for a budget line.

AI is the newest ERG

This is the part that sounds strange the first time you hear it. AI is not a person. AI is not a population. So how can it be an ERG.

It can be an ERG because of how it is showing up, not because of what it is.

AI is showing up in your company without a clear invitation. The system does not know what to do with it. People are reacting to it the same way companies once reacted to veterans, to working moms, to Black employees, to disabled employees. With confusion. With fear. With overcorrection in some places and total avoidance in others.

AI is a new member of the team. Not a tool you install.

Treating AI like an ERG does not mean throwing it a heritage month. It means using the same playbook ERGs have always used to integrate something new and produce real value from it.

Start with belonging. Decide where this new thing actually fits. Give it a clear role. Decide what it is allowed to do and what it is not. Decide who speaks for it when something goes wrong.

Then work on being seen. Have honest conversations about what people actually think of it. Make space for the people who are scared of it. Make space for the people who are excited about it. Stop pretending everyone feels the same way.

Then build trust over time, the way every successful ERG has built trust over time. By producing something the company could not get any other way. AI will earn its place in your company the same way every group before it has. By making the company better at something that mattered.

Why this is good news for ERGs, not just for AI

Here is the part that should land for every ERG leader, every DEI professional, and every employee who has ever sat in an ERG meeting wondering if anyone was paying attention.

When AI gets treated like an ERG, every existing ERG in the company becomes more important, not less.

Five reasons why.

1. ERGs stop being a side project. For years, ERGs have had to fight for their budget and explain why they matter. The moment AI is treated like an ERG, ERGs are doing the same kind of work the CEO is losing sleep over. The conversation changes from what events did you run to what are you teaching us about how to bring something new into this company.

2. What ERGs already know becomes valuable to everyone. Veterans groups know how to help someone adjust when their old world spoke a different language. Disability groups know how to make accommodation feel normal instead of awkward. Working parents groups know how to redesign work so a new reality can fit. Every one of those skills is exactly what AI integration needs.

3. ERG leaders get a real seat at the table. Right now the AI conversation lives with IT, Legal, and the executive team. ERG leaders are usually nowhere near it. If AI is really an integration challenge, then ERG leaders are some of the most qualified people in the building to weigh in. They should be in those rooms.

4. ERGs become recognized as the part of the company that knows how to make new things belong. That is a real and rare skill. Most companies are not good at it. The companies that are good at it are good because of their ERGs, even if they have never said it out loud.

5. ERGs become harder to cut. ERGs are usually the first thing on the chopping block when budgets get tight or when political winds shift. The moment they are recognized as the engine that helps the company integrate AI, and whatever comes after AI, they stop being optional. You do not cut the function that is helping you handle the biggest change in front of the business.

What companies can actually do, starting now

This does not have to be a giant project. There are simple things any company can do in the next few months that would change how this lands.

1. Stop calling AI a tool problem. In the next AI meeting, ask a different question. Not what tools are we using. Ask how are we bringing this new thing into the company, and who is good at that kind of work. See who in the room has answers and who does not.

2. Get the ERG leaders in a room together. Bring them together and ask them straight up. What do you know about helping new people belong here. What do you know about what makes integration succeed and what makes it fail. Take notes. That conversation is gold.

3. Put ERG leaders on the AI team. Not as guests. Not as a check the box move. As actual contributors. Put them on AI committees, ethics groups, and rollout teams. They are not there to represent their group. They are there because they know how to integrate.

4. Look at what each ERG already knows and write it down. Veterans groups know cross system integration. Disability groups know accommodation as a default. Working parents groups know how to redesign work around a new reality. Pull those lessons out of people's heads and into a real document the AI team can use.

5. Move ERGs out of the culture budget and into the strategic budget. If ERGs are doing strategic work, fund them like strategic work. Change who they report to. Change how they are measured. Change how seriously they are taken in the room.

6. Use what you build now for what comes next. AI is not the last big change that will hit your company. There will be another one. And another one after that. Building the ability to integrate something new is one of the most valuable things a company can have right now.

The choice in front of every company

There are basically two paths.

The first path. Keep treating AI like a software rollout. Keep treating ERGs like a culture program. Watch the AI work stall in the same uneven way it is stalling now. Watch the ERG budget get cut the next time the company tightens up. Three years from now, look around and notice that another company figured this out and you did not.

The second path. Treat AI like a new member of the team. Bring it in through the part of the company that already knows how to do that. Give the ERG leaders real authority. Take what they have learned over forty years of doing quiet, hard, mostly thankless work, and make it the engine that helps the rest of the company adapt to whatever comes next.

Things are changing fast right now. They are going to keep changing. The companies that handle the change well are not going to be the ones with the biggest tech budget. They are going to be the ones who already know how to help people, and now AI, find a place inside the company.

Companies already have those people.

They are the ERGs. And it is time for everyone to start treating them that way.

We build the architecture for what most change models skip: the human.

Sara B. McNamara, MSOD Founder, HumanWorks™ Solutions