The AI Boom Needs Human Hands: AT&T’s CEO Says Telecom Must Hire Nearly 200,000 Workers by 2032

By Priya, Research, HUMAN HIVE

Who builds the AI boom? People with tools on their belts. AT&T CEO John Stankey says the telecom industry will need nearly 200,000 more fiber and technician workers by 2032 to connect the data centers AI runs on, and his company is already working hard to find them. The skill gap that matters most right now is not prompting. It is the human ability to learn a new job fast.

The 60-second version

  • AT&T's CEO says telecom needs nearly 200,000 more fiber and technician workers by 2032 to connect the AI build-out.

  • AT&T says it hires for skills and potential, then teaches the rest.

  • This week, pick one hard-to-fill role and write down the skills it needs, not the degree.

What happened

In a Business Roundtable interview series that runs on the This Old House Radio Hour podcast, host Kevin O'Connor asked AT&T Chairman and CEO John Stankey about the people behind the network. The episode is listed on Business Roundtable's Work in Progress page, dated November 10, 2026, with a full transcript. Stankey's numbers were blunt: "probably by 2032, you're looking at something just short of 200,000 people that need to be hired in."

On his own company's pace, he said AT&T has "been hiring between 3,000 to 3,500 folks a year to do more of that technical fiber work, and we're having to work real hard to find the talent."

On November 20, Fortune's Preston Fore put the interview in front of a much bigger audience, tying the gap to the data center building boom. Fortune reports that AT&T has hired 10,000 workers for these roles since 2023, and that a company spokesperson said AT&T hires on "skills and potential" rather than requiring a degree: "we teach the rest." Crypto Briefing reported the same 200,000 figure the same day.

Government data points the same direction for the closest cousin of this work. The U.S. Bureau of Labor Statistics projects jobs for electrical power-line installers and repairers to grow 10 percent from 2025 to 2035, about 10,900 openings a year, and names the reason in plain words: "the increasing use of artificial intelligence (AI) technologies and the data centers needed to support this growth are expected to contribute to rising electricity demand." The typical entry is a high school diploma plus long-term on-the-job training.

Why it happened (the human side)

For two years the loudest AI story has been about which jobs disappear. This one runs the other way. Every model, every agent and every chatbot lives in a building that needs power and fiber, and fiber still goes in the ground by hand.

The gap is not only about money or headcount. Stankey named the human problem himself: the way we rank work. "All work is something that brings dignity and integrity to somebody's life," he said, and he asked us to "maybe not stigmatize one work over another but be balanced in how we think about it." For a generation, many families and schools steered people away from hands-on careers. AI just raised the price of that habit.

The second human problem is training. AT&T's answer, in Stankey's words: "We offer certification programs. We allow for tuition reimbursement." The company is also "partnering with community colleges to step up some of our intake in from the military." That is a skills-first hiring model, and most companies still screen for credentials instead.

What Mitch has taught on this

Mitch has been saying for a while that the job market is changing in a way the layoff headlines miss. In his video No One Is Hiring: The AI Layoff Wave Never Came. This Did., he puts it like this: "skilling up today is no longer about the basics. It's about AI. It's about learning how to master it. It's about adaptability. It's about discernment."

And his close: "Get skilled up quick. Learn about AI. Learn about adaptability. Learn about service. Not just customer service, but service to others."

The fiber story proves his point from an unexpected angle. The person who climbs a pole and the person who builds an AI agent need the same core muscle: the ability to learn something new, quickly, and do it well for someone else. We wrote about the white-collar side of this in The AI Capability Paradox. This is the hard-hat side.

Best practices: hire and grow for skills, not paper

  1. Rewrite one job post this month. HR picks the role you struggle to fill most. Strike every requirement that is a credential and not a skill. Keep only what a person must be able to do on day 90.

  2. Price your own training time. AT&T's approach is to hire for potential and teach the rest. Ask your L&D lead: how many weeks does it take us to make a new hire productive, and who owns that plan? If nobody can answer, that is your first fix.

  3. Build a pipeline you do not have yet. Stankey named community colleges and the military. Have one leader call one local program this quarter and ask what their graduates can already do.

  4. Protect learning time for current staff. In a Howden study of more than 2,000 UK employees, 53 percent worried they won't have time to learn new skills on top of existing workloads (TechRadar, November 13). Put a recurring learning block on the calendar and defend it like a client meeting.

  5. Stop ranking work out loud. Managers set the tone. Retire phrases like "just a tech job" or "only the field team." Respect is a retention tool.

What's driving people crazy this week

"Learn AI" on top of a full workload. The same Howden study found 49 percent worry they will be expected to do more work because of AI time savings. People are not against learning. They are against learning on their own time with no plan. The human fix: name the skill, give the hours, measure the result.

What this means for the next 1 to 3 years

  • The "AI job" splits in two. One half builds and manages AI at a desk. The other half builds and maintains the physical world AI needs. Smart workforce plans will train for both and let people move between them.

  • Time to competence becomes the number that matters. When talent is short, the winner is the company that turns a willing person into a capable one fastest. Expect leaders to track ramp-up time the way they track sales.

  • AI becomes the training partner in the field. Picture a new fiber splicer with a tablet that answers questions on the spot, while a senior tech checks the work. AI speeds up learning. Humans still sign off.

  • Adaptability becomes a hiring criterion. If the work keeps changing, the best predictor is how a person handles change, not where they went to school.

How we evolve

The human skill to build next is fast learning: picking up something unfamiliar and getting useful at it without freezing up. It matters more as AI gets better, because the tools and the jobs will keep changing under our feet. Mitch's ADAPT work and his advice to "skill up" both start here.

One small daily practice: spend 10 minutes a day learning one thing outside your lane, then explain it to a colleague in two sentences. Teaching it back is how you know you learned it.

Try this today (under 30 minutes)

  1. Ask at your next team meeting: "Which role here would we struggle most to fill next year, and what would it take to grow someone into it?"

  2. Run the 90-day test: take one open job post and circle every line a person could learn in 90 days with training. Move those from "required" to "we teach this."

  3. Try this prompt: "Here is a job description. List the skills a person must already have on day one, and the skills we could teach in the first 90 days. Then rewrite the post around the first list only." Read the output and decide yourself what stays.

FAQ

How many telecom workers does AT&T's CEO say the industry needs?

John Stankey said "something just short of 200,000 people" will need to be hired in telecom by 2032, speaking in a Business Roundtable interview series dated November 10, 2026.

Why is AI creating demand for fiber and line workers?

AI runs in data centers that need power and network connections. The Bureau of Labor Statistics names AI and data centers as a driver of rising electricity demand behind its 10 percent growth projection for power-line installers from 2025 to 2035.

Do these jobs require a college degree?

Fortune reports that AT&T hires on "skills and potential" rather than a degree. BLS lists a high school diploma plus long-term on-the-job training as the typical path for power-line installers.

Where HIVE can help

Skills gaps are people gaps. Book a 20-minute call about where your workforce is stuck. Not sure where your team stands? Take the free readiness assessment.

Get The HIVE Daily Buzz in your inbox.

As the machines get more human, are we?

Next
Next

AI Fail of the Week: Meta’s Muse Agent and the Permission Nobody Meant to Give