Written by Gini Graham Scott, PhD.
In 2026, the fastest way to see how AI is changing work in the United States is to walk through a few offices, warehouses, studios, and control rooms and listen to what people are actually doing. The “AI job market” is a loose federation of about fifty emerging roles that cluster into seven broad families — from the engineers wiring up giant models to the educators and coaches trying to help everyone else keep up.
The most visible group are the builders, which include AI engineers, machine learning engineers, infrastructure engineers, GPU (Graphics Processing Unit), cluster managers, solutions architects, integration specialists, workflow architects, robotics techs, and autonomous systems supervisors. These are the people who are constructing and wiring the systems that now sit underneath banking, logistics, media, and even parts of government. In the U.S., AIrelated technical roles have grown into the hundreds of thousands, with AI engineer and ML (machine learning) engineer among the top growth titles in tech job postings over the last few years. These jobs remain concentrated in some of the largest cities, including San Jose, San Francisco, Seattle, New York, Boston, Austin, but they are beginning to spill into emerging hubs like Raleigh, Houston, and Denver as data centers and robotics corridors expand.
Demographically, those holding these jobs look a lot like the rest of the high end tech world: skewed toward younger workers and toward men. But the field is slowly diversifying as more women and underrepresented groups move through computer science and data science programs. Typical entry paths are getting a bachelor’s or master’s degrees in computer sciences, statistics, or engineering; internships at big tech firms; side projects and Kaggle competitions, which are challenges for competitors at the different stages of their machine learning careers. However, there are a growing number of other ways to get started, such as bootcamps, employer-led reskilling programs, and “second degree” moves from traditional software roles into AI infrastructure development, called MLOps (Machine Learning Model Operationalization Management) which involves a machine learning development process to design, build and manage reproducible, testable, and evolvable ML-powered software. Such ways to learn the new skills required by AI jobs, particularly for workers in their 30s who are eager to become part of the AI revolution.
While the builders write the code, another major source of jobs is the people who teach and tame the machines. These jobs include AI trainers, data annotators, prompt engineers, generative AI specialists, conversation designers, humanAI interaction designers, and knowledge architects. These are the teachers who makes abstract models useful, and there are now tens of thousands of these roles worldwide. Some of these workers are early-career humanities graduates or gig workers doing labeling and safety rating from spare bedrooms. Others are midcareer UX (User Experience) designers who create products, systems, or services that provide meaningful, intuitive, and enjoyable experiences for users. Still other jobs in this category of AI teaching and design include service designers, and technical communicators who have evolved into prompt engineers and conversation designers inside large companies.
This cluster is more mixed by age and gender than the pure engineering track, with a substantial number of women in design and interaction roles and many workers in their 30s and 40s pivoting into this type of AI work from customer support, marketing, or operations. Their training is based on short learning modules, such as taking short courses in prompt design, UX for conversational interfaces, AI fundamentals; going to internal bootcamps that teach employees how to rewrite workflows around AI tools; and obtaining online certificates in applied machine learning. While the biggest contracts are still signed in San Francisco and New York, much of the work is geographically diffuse, with remote contractors spread through lowercost metros and college towns, especially for annotation and RLHFstyle training, which refers to reinforcement learning from human feedback, a technique to align an intelligent agent with human preferences.
As AI systems have played an increasing role in consequential decisions that reverberate through society, a third type of field has been growing: the guardians, who are designed to establish and implement controls. These jobs include AI ethicists, governance specialists, safety researchers, compliance officers, auditors, policy advisors, legal specialists, red team specialists, cybersecurity analysts, and deepfake detection analysts. Their role is to try to keep the systems safe, fair, and legal. Many of these jobs sit in highly regulated industries, such as in hospital systems, insurers, and federal and state agencies in Washington and state capitals.
The demographics for these jobs are quite different. Rather than the twentysomethings who are becoming coders and taking other technical jobs, the job holders are more likely to have a background as a former civil rights lawyer, midcareer risk manager, a cybersecurity veteran, or policy analyst who once focused on privacy or discrimination law. Women are better represented in this type of job than in core engineering, especially in policy, ethics, and legal roles. The training pattern is equally hybrid. Some common backgrounds include a JD plus an AI law course; a background in internal audits plus a responsibleAI certification; a PhD in philosophy or social science combined with technical reading groups and industry fellowships. As governments roll out new AI rules and companies face real regulatory pressure, the demand for these skills has been climbing faster than traditional computer science departments can adapt.
The fourth type of job is where AI combines with art and media. Such jobs include AI content editors, factcheckers, video producers, image designers, music producers, synthetic voice designers, and avatar creators. Their jobs involve turning raw model output into stories, visuals, sounds, and characters that real audiences will accept. In newsrooms and media startups, AI content editors and factcheckers play a growing role in making sure that the increasing use of AI doesn’t strain credibility, cleaning up drafts, and making sure that machinegenerated copy doesn’t lead to mistakes in facts when published. Increasingly, in film studios and ad agencies in Los Angeles and New York, AI video producers and image designers use models as idea generators and timesavers, while still relying on years of human craft to decide what is good enough to produce.
Unlike the engineering type of job, this creative labor market draws heavily from existing pools, such as journalists in their 30s and 40s, designers and editors, audio engineers, and motion graphics artists. Many already survived one wave of digital disruption, and now they are wary but pragmatic about this one. Training tends to be self-directed and intensely practical, such as learning to more skillfully use prompts and control the specific tools a newsroom, brand, or production company uses, rather than going back to school for formal degrees. The work remains concentrated in legacy media and entertainment hubs, but remote work has pulled in people from far beyond New York and Los Angeles, especially in editing, factchecking, and motion graphics, which are roles that can be done from anywhere with a decent connection.
Running parallel to these creative jobs is a fifth cluster — the business translators. This category includes AI product managers, marketing strategists, advertising creatives, sales enablement specialists, customer success managers, transformation leads, automation consultants. Their role is turning the use of AI into actual revenue, cost savings, and performance metrics. Across 2024 and 2025, AIrelated responsibilities began appearing in a rising share of product and strategy postings, and by 2025 roughly 4–5% of all U.S. job ads included usingAI. Inside companies, that usage included humans deciding which workflows to automate, which features to launch, and how to reskill existing staff.
These workers are overwhelmingly midcareer. It is common to meet a 38yearold product manager who started in mobile apps, a 45yearold operations director who has become an AI transformation lead for a bank, or a marketer who now wears the title “AI marketing strategist.” Gender representation is closer to balance, and backgrounds are varied, including such past roles as MBA graduates, former consultants, sales leaders, and data-driven marketers. Their training pathways are executive-oriented, such as participating in short “AI for business” programs, inhouse academies, and on-the-job experimentation with analytics and generative tools. Geographically, they are based where corporate headquarters are located, such as New York, Chicago, Dallas, Atlanta, and San Francisco, but due to remote and hybrid policies, the same job holders might show up in smaller cities, such as Boise or Cleveland, creating the potential for a widely distributed team.
The sixth category of jobs is a reminder that AI is not just a tech story. Rather AI is embedding itself everywhere. Under titles like AI healthcare analyst, clinical AI coordinator, drug discovery scientist, AI financial analyst, algorithmic trading specialist, AI is threading itself into medicine, pharma, finance, and beyond. For example, in big academic hospitals, clinicians and analysts work together to interpret risk scores and alert systems instead of just relying on static protocols. In biotech companies and pharma R&D labs, AIdriven scientists narrow down the molecules to be investigated long before expensive lab work begins. On trading floors and in backoffice risk teams, AI-augmented analysts watch for patterns that humans would struggle to spot on their own.
Here, the workforce comes almost entirely from domain expertise. A clinical AI coordinator is usually a nurse, pharmacist, or physician who has added informatics and AI literacy to their work performance. A drug discovery scientist was already a chemist or biologist. An algorithmic trader probably started as a quant, which is a financial professional who uses advanced mathematics, statistics, and computer programming to analyze financial data, manage risk, and develop automated trading strategies. These roles cluster in the industry’s traditional hubs, such as Boston and Houston for healthcare, Boston, Cambridge, and San Diego for biotech, New York and Chicago for finance. At the same time, these jobs using AI are also beginning to appear in regional hospitals and regional banks that license AI tools instead of building them inhouse.
Finally, there is a seventh, often overlooked category of AI jobs, the guides and overseers. Such jobs include AI education specialists, curriculum developers, career coaches, human oversight specialists, and search optimization specialists — jobs which exist to provide help in working with and using AI. For example, in community colleges, AI curriculum developers rebuild syllabi so students learn how to use AI without abandoning critical thinking. In workforce centers, AIliterate career coaches sit with laidoff workers from retail or call centers and map realistic transitions into AI-adjacent roles, from customer success to frontline oversight. Inside companies, human oversight specialists monitor fleets of AI agents, stepping in when systems fail and documenting those failures so models and policies can be adjusted.
This is the most demographically diverse cluster of all. It includes K–12 teachers in their 50s, instructional designers in their 30s, counselors and social service workers in their 40s, and younger digital native coaches who’ve made themselves fluent in job platforms and AI tools. Women are strongly represented, as they are in education and counseling more broadly. Training comes from a mix of traditional education degrees, new AI literacy certificates, and employer-sponsored programs. And geographically, these jobs are everywhere. They’re in any state that invests in workforce transition, any district experimenting with AI in classrooms, and any large employer rolling out AI at scale needs people who can translate, coach, and watch.
In sum, in the last few years, the U.S. has added about a hundred thousand AIrelated roles per year, with millions of existing jobs absorbing AI responsibilities as part of their evolution. The geography is uneven: San Jose and San Francisco still build the bulk of the systems, New York dominates AI/intensive finance and media, and a handful of metros, such as Austin and Raleigh, combine relatively high AI wages with lower costs of living. The demographics are uneven as well, with white males tending to dominate in technical roles and more diversity in governance, education, and support.
But the increasing trend towards more and more AI jobs is clear, with dozens of distinct roles spread across seven broad categories, touching everything from robot maintenance and deepfake detection to career counseling. Some of these jobs are glamorous; some are invisible; some are stopgaps that may not survive the decade. Together, though, they describe a labor market that is being quietly rebuilt due to the growing presence of intelligent systems, along with the humans who design them, question them, supervise them, and teach the rest of us how to live with them.
____________
Gini Graham Scott, Ph.D. is the author of over 50 books with major publishers and has published 200 books through her company Changemakers Publishing and Writing (http://www.changemakerspublishingandwriting.com). She writes books, proposals, and film scripts for clients, and has written and produced 20 feature films and documentaries, including Conned: A True Story and Con Artists Unveiled¸ distributed by Gravitas Ventures. (http://www.changemakersproductionsfilms.com).
Her latest books include Ghost Story and How to Find and Work with a Good Ghostwriter published by Waterside Productions; and The Big Con and I Was Scammed, published by American Leadership Press. She additionally has designed and published over 100 games with ALB Games.
