Career Pathways in AI — Security Medic Consulting

Students & Early Career / Career Pathways

12th Grade · First-Year College · Job Seekers

Careers in AI aren’t just for coders. The real map is wider, weirder, and much more hopeful than the headlines suggest. Here’s the honest picture of what roles exist, what to study, and how people actually break in.

§01 The map nobody drew for you

“A career in AI” doesn’t mean what you think it means.

When someone says “a career in AI,” most people picture a narrow image: a twenty-something in a hoodie training neural networks in a quiet room lit only by a monitor. That image is real — for maybe half a percent of the people who will build AI-shaped careers over the next decade. The other 99.5% are doing something different.

They’re teachers integrating AI into how students learn. Nurses using AI to triage patients. Lawyers who became the AI-policy expert at their firm. Marketing managers who got good with AI tools and quietly became their team’s secret weapon. Police chiefs learning how criminals use AI so they can defend against it. Farmers using AI for crop monitoring. Insurance adjusters. Accountants. City planners. Product managers. Graphic designers. Small business owners. High school principals.

These are all AI careers. None of them require a PhD in machine learning. Some of them don’t require any coding at all. What they all require is the same thing: the ability to think clearly about what AI does well, what it fails at, and how to apply it responsibly in a specific domain you understand.

The narrow definition of “AI career” is the reason so many thoughtful young people wrongly conclude the field isn’t for them. This piece is here to correct that picture.

The reframe

AI is becoming like electricity. A hundred years ago, working with electricity meant you were an electrical engineer. Today, almost every profession uses electricity — the engineers are still there, but they’re the minority. The majority of electricity-enabled careers belong to everyone else.

AI is going through the same transition, faster. The builders exist, and they matter. But the much larger opportunity is for people who work with AI — not inside it — in every other field that exists.

§02 Four broad paths into AI-shaped careers

Pick the path that matches who you actually are.

Careers in and around AI fall into roughly four categories. Most people don’t know about more than one of them. Once you see all four, you’ll recognize which one (or which combination) fits you.

Path 1 — Build the AI (the builders)

These are the people creating, training, and fine-tuning the models themselves. Machine learning engineers, research scientists, AI infrastructure engineers. They work at companies like Anthropic, OpenAI, Google DeepMind, Meta AI Research, Microsoft Research, and hundreds of smaller labs and startups.

This path requires deep technical skills: strong math, serious programming, and usually a bachelor’s or graduate degree in computer science, math, physics, or a related quantitative field. It’s rewarding, well-paid, and extremely competitive. Maybe 5% of AI-adjacent careers are here.

Path 2 — Deploy the AI (the implementers)

This is the much larger group — the people taking existing AI tools and putting them to work inside organizations. AI engineers (different from ML engineers; they integrate rather than train), AI product managers, solutions architects, AI consultants, developer advocates, AI operations specialists. They bridge between the builders and the users.

This path can be entered with a bachelor’s in computer science, engineering, business, or any related field — but increasingly also through bootcamps, certifications, and real project experience. Coding skills help but aren’t strictly required for all of these roles. Maybe 20% of AI-adjacent careers are here, and this share is growing fast.

Path 3 — Apply the AI in a domain (the practitioners)

This is where most of the growth is happening, and where most young people don’t realize the opportunity lives. These are the nurses, teachers, lawyers, social workers, designers, marketers, accountants, police officers, farmers, and small business operators who become the AI-expert in their domain. They don’t build or even deploy AI — they’re the people who know how to use it well in their specific field, and who understand the limits.

You get here by being genuinely good at a field you care about, and then being the person who leans into AI early. The credential is your domain expertise plus your fluency. Formal AI training helps, but experience and reputation matter more. Maybe 70% of the AI economy lives here. It is the most overlooked category.

Path 4 — Govern the AI (the stewards)

These are the people making sure AI is used safely, ethically, and legally. AI ethicists, AI policy analysts, compliance officers, AI auditors, risk managers, privacy specialists, AI lawyers, academic researchers studying AI’s social impact. The field didn’t really exist five years ago and now it’s hiring fast.

Paths in: law, public policy, philosophy, ethics, social sciences, journalism, regulatory compliance, cybersecurity. You don’t need to know how to build an LLM, but you do need to understand what one does, where it fails, and what harm looks like. Maybe 5% of the current AI economy, with faster growth ahead as regulation catches up.

§03 Specific roles, in plain terms

Eight real roles that exist right now — including the ones that don’t require code.

These are representative, not exhaustive. Each card tells you what the job is, what an entry-level version looks like, and the shortest honest path in.

Path 1 — Build the AI

Machine Learning Engineer

Builds, trains, and improves the AI models themselves. Writes code that turns raw data and model architectures into working systems. Works closely with research scientists on the frontier, or on applied teams that fine-tune existing models for specific use cases.

Entry point

Junior ML engineer or ML ops engineer at a tech company

Typical path in

Bachelor’s in CS, math, or physics; strong portfolio; internships

Coding required

Yes — Python, deeply

Who this suits

Students who genuinely enjoy math and programming, and who find the technical details of how models work intrinsically interesting. If you’ve ever wanted to read a research paper just to understand how something works, this might be you.

Reality check

Highly competitive. The bar is rising — many companies now prefer master’s or PhD candidates for research roles. Entry-level applied ML engineering is still accessible with a strong bachelor’s and good projects, but expect to work hard to stand out.

Path 2 — Deploy the AI

AI Product Manager

Decides what AI-powered products or features get built, by whom, and for what users. Works with engineers, designers, and executives to turn a vague “we should use AI for this” into a shipped product. Increasingly one of the hottest roles in tech.

Entry point

Associate PM at a tech company; often internal transfers from support or operations

Typical path in

Bachelor’s in anything + technical fluency + product sense

Coding required

Helpful, not required

Who this suits

People who like thinking about users, tradeoffs, and what good products feel like. You need to understand AI well enough to have sharp opinions about what it can and can’t do, but you don’t need to write the code. Strong writers and communicators thrive here.

Reality check

PM roles often look like they require experience you don’t have yet. The honest entry route is to either (a) become a PM at a smaller company first, or (b) join a larger company in another role — support, operations, marketing — and transfer into product when an AI PM role opens internally.

Path 2 — Deploy the AI

Solutions Engineer / AI Solutions Architect

Works with companies that are adopting AI, figuring out how to actually make it work in their specific environment. Part engineer, part consultant, part teacher. Might help a hospital integrate AI transcription, help a law firm set up a document review system, or help a bank deploy a compliance chatbot.

Entry point

Junior solutions engineer, implementation specialist, or customer engineer

Typical path in

Bachelor’s in CS, engineering, or technical field; comfort with code and customers

Coding required

Yes — but applied, not research-level

Who this suits

People who like both technical problems and humans. If you enjoy explaining things, like variety, and get energized by solving a new puzzle each week, this is one of the most underrated roles in AI. It’s also one of the most direct routes to learning which industries pay well for AI expertise.

Reality check

Travel is often part of the job. Expect to spend meaningful time in front of customers. If you want to sit quietly and write code alone, this isn’t your role — it’s the opposite.

Path 3 — Apply the AI

AI-Fluent Domain Expert (Marketing, Sales, HR, Ops, etc.)

You pick any field you want — marketing, sales, human resources, operations, finance, accounting, supply chain — and become the person on your team who’s genuinely good at applying AI to that field’s problems. Not a new job title, necessarily. But the people doing it are getting promoted and paid faster than their peers.

Entry point

Any entry-level role in the domain you care about

Typical path in

Bachelor’s in your chosen field; bring AI fluency as your edge

Coding required

Who this suits

Almost anyone. This is the single most accessible path into AI-adjacent work, and the one most overlooked by 18-year-olds who assume they have to major in computer science. If you know you like marketing or accounting or operations, major in that — and then aggressively learn to use AI tools in that context. You’ll be rare and valuable.

Reality check

“Learning to use AI in X” sounds vague because it is. The way it becomes concrete is through projects: a side project where you automate something in your field, a portfolio where you document what you learned, a presentation to your team. Make the fluency visible, or no one will know you have it.

Path 3 — Apply the AI

Healthcare + AI (Nursing, Medical Tech, Clinical Informatics)

Healthcare is one of the largest and most AI-impacted domains. Nurses and clinicians who understand AI tools are in enormous demand. Clinical informatics — a field combining healthcare and IT — is growing fast and pays well. Hospitals need people who can bridge the gap between doctors and technology.

Entry point

RN, medical tech, or clinical informatics associate

Typical path in

Nursing degree (ADN or BSN), medical tech certification, or informatics program

Coding required

No (for clinical roles); some (for informatics)

Who this suits

Students drawn to healthcare who also care about technology. The stereotype that nursing is “not a tech job” is outdated — the best nurses are increasingly AI-fluent, and that skill gets them leadership roles, higher pay, and policy-making positions.

Reality check

Healthcare is heavily regulated. Entering through a traditional healthcare path (nursing, tech, allied health) and layering AI fluency on top is more reliable than trying to enter as a “healthcare AI specialist” from the tech side. Domain credentials matter here.

Path 3 — Apply the AI

Education + AI (Teachers, Curriculum Designers, EdTech)

Every school district in the country is wrestling with how to integrate AI into teaching and learning. Teachers who can thoughtfully use AI — not just permit it — are in short supply. Curriculum designers who build AI-aware materials are getting hired. EdTech companies are paying real money for former teachers who understand classrooms.

Entry point

Teacher, curriculum designer, or EdTech junior role

Typical path in

Education degree or teaching certification; AI fluency as differentiator

Coding required

Who this suits

Students who love teaching and worry the profession is being “automated away.” It’s not. The teachers being valued most right now are the ones who understand AI well enough to integrate it intelligently. If you’ve always wanted to teach, this is one of the best moments in a generation to enter the field.

Reality check

Teaching pay still varies wildly by region and district. But the AI-fluent teacher has significantly more career options now — instructional design, EdTech, school leadership, consulting — than the same teacher would have had ten years ago.

Path 4 — Govern the AI

AI Policy, Ethics, and Governance

The people thinking about what AI should and shouldn’t do, how to regulate it, how companies should use it safely, and what rights individuals have around it. Works in government agencies, law firms, think tanks, universities, and inside companies as policy or ethics leads.

Entry point

Policy analyst, compliance associate, research assistant, junior legal role

Typical path in

Bachelor’s in public policy, law, philosophy, or social sciences

Coding required

Who this suits

Students drawn to big questions about technology, power, and society. If you’re good at writing, research, and thinking about systems — and if you’re skeptical of pure-hype takes on AI — this field needs you. It’s one of the fastest-growing AI-adjacent areas and one of the most underserved.

Reality check

Many of these roles are in government, nonprofits, or academia, which pay less than tech. The exception is AI governance roles inside large companies, which pay well but are harder to enter. Graduate school (law, policy, or a masters in technology ethics) is common for leadership roles.

Path 4 — Govern the AI

AI Security and Risk (Cybersecurity + AI)

Specialists who defend AI systems from attack, and who use AI to defend other systems. One of the fastest-growing fields in cybersecurity, with a severe shortage of qualified people. Works in corporations, government, critical infrastructure, and security consulting firms.

Entry point

SOC analyst, security engineer, or threat intelligence analyst

Typical path in

Bachelor’s in cybersecurity, CS, or IT; certifications (Security+, CISSP path)

Coding required

Helpful — scripting more than full engineering

Who this suits

Students drawn to the cat-and-mouse dynamics of security — understanding how systems break, and how to defend them. Cybersecurity has been short-staffed for a decade; adding AI expertise makes you even rarer. This is one of the most employable paths on this list.

Reality check

Entry-level cybersecurity is getting harder because AI is handling more tier-1 analyst work. The way in is through internships, home labs, certifications, and genuinely learning the fundamentals. This is a field where self-teaching and demonstrated skill matter as much as formal credentials.

You don’t have to love AI to build a career around it. You have to be willing to understand it, work with it, and bring something else — a field, a skill, a perspective — that it can’t replace.

§04 What to study — and the myth of the “right” major

The major matters less than you’ve been told. The habits matter more.

Students get asked what they want to major in before most of them have had a single conversation with someone who works in a field they’re considering. The pressure to “pick right” is real. But the honest truth is that the specific major matters less now than it used to — and less than your parents and counselors probably think.

Here’s what the evidence actually shows about AI-era careers:

What matters more than your major

The ability to write clearly, think critically, work in teams, explain things to non-experts, learn new skills fast, and demonstrate what you’ve built. These are durable across almost every AI-shaped career. Employers repeatedly report that they care about these more than any specific credential.

Good majors for AI-adjacent careers

This is not a ranked list — it’s a map. Each of these has paths into AI careers.

Computer science — direct entry to builder and deployer paths. Still valuable; pick it because you love it, not because it’s “safe.”
Data science / statistics — opens up ML engineering, data engineering, analytics, and many practitioner roles.
Cybersecurity — one of the best-employment-rate fields on this list, with AI expertise layered on top.
Engineering (any kind) — strong problem-solving foundation, opens up both deployer and practitioner paths.
Mathematics / physics — deep technical path; often leads to graduate school and research roles.
Nursing / healthcare — one of the most AI-resistant and well-compensated domain paths.
Education — strong domain path; high demand for AI-fluent teachers and curriculum designers.
Business / management — strong base for PM, operations, consulting, and business-side AI roles.
Communications / writing — underrated; strong writers and communicators are in demand everywhere AI is being deployed.
Psychology / cognitive science — valuable for UX, AI ethics, human-computer interaction, and product research.
Public policy / political science / law — core path to AI governance and policy work.
Philosophy / ethics — surprisingly valuable for AI ethics, alignment research, and policy.
Design (graphic, UX, product) — AI is reshaping design tools; designers who lean in are in demand.
Social sciences (sociology, anthropology, economics) — valuable for AI impact research and policy work.

Notice what isn’t here: there’s no specific “AI major.” Some universities are launching them, but they’re new and the jury is still out. For most students, a solid major in a field they enjoy, plus deliberate AI fluency on the side, is a stronger position than a brand-new degree program.

Paths that don’t require a four-year degree

This is underdiscussed. Many high-paying AI-adjacent roles can be entered without a bachelor’s degree:

Skilled trades (electrician, HVAC, plumbing) — not AI careers in the narrow sense, but extremely AI-resistant and in high demand. Increasingly well-paid.
Coding bootcamps (12-24 weeks) — can lead to junior developer and implementation roles at companies willing to hire for skill rather than credential.
Cybersecurity certifications (Security+, Network+, then stacking) — a well-traveled path into security operations roles.
Community college programs in IT, cybersecurity, healthcare tech, or data analytics — often two-year programs with strong employment outcomes and low debt.
Apprenticeships in tech (still rare in the U.S., but growing) — earn while you learn programs from major employers.
Self-taught with portfolio — harder path, but real: build things, put them online, apply widely. Works best in software, data, and design.

A note about student debt

A four-year degree is a powerful thing and also an expensive one. The value proposition is real — but it’s not universal, and it’s not without cost. If you’re weighing a four-year school against a community college, a trade, or a specialized program, do the math honestly. A bachelor’s with $80,000 in debt and an unclear career path is not automatically better than a trade apprenticeship that starts paying at 18. The right answer depends on you, your field, and your finances — not on what anyone tells you is “the right thing to do.”

§05 How people actually break into AI careers

The real patterns, not the LinkedIn versions.

If you read LinkedIn posts from successful AI professionals, the path sounds linear: went to the right school, got the right internship, took the right first job, now runs an AI team. That’s the retrospective story. The real path usually looks a lot messier.

Here are the patterns that keep showing up among people who actually have AI careers today.

Pattern 1 — They showed up before they were “ready”

They joined AI clubs, took online courses before their curriculum required them, built side projects that had nothing to do with their classes, and volunteered for AI-related work at their first jobs. They didn’t wait until they felt qualified. Feeling qualified came from doing, not the other way around.

Pattern 2 — They built something public

A small project, a blog, a GitHub repo, a short paper, a presentation to a student group. Something specific that another person could look at. “I’m interested in AI” is an invisible claim. “Here’s a tool I built that helps students plan study schedules using AI” is a visible one. The public artifact is often the difference between getting an interview and not.

Pattern 3 — They said yes to the AI stuff at their first job

Almost everyone I know with a serious AI-era career got started by being the “AI-curious person” at a job that wasn’t originally about AI. The company needed someone to explore how AI might help the team. They raised their hand. Six months later, they were the AI lead. That pattern keeps repeating.

Pattern 4 — They networked earlier and more consistently than they expected

Talking to people in fields you’re considering is the single highest-return activity for a student. Most people never do it, or do it badly. The ones with AI careers tend to have reached out to strangers on LinkedIn, attended talks, emailed professors, asked professionals for 20-minute conversations — and showed up prepared. If you can force yourself past the discomfort, you’ll learn more in five of those conversations than in six months of scrolling career pages.

Pattern 5 — They kept learning after school ended

AI moves too fast for any degree to keep up. The people who thrive are the ones who never stop learning: new tools, new techniques, new applications. Not frantically — just consistently. An hour a week, forever, on something that interests them. The compounding effect is massive.

§06 Myths that keep people out

Things you might believe that are holding you back.

AI career myths, honestly corrected

Myth “You have to be a math genius to work in AI.”

Reality Some AI roles — research, ML engineering — require strong math. Most don’t. Product managers, policy analysts, solutions engineers, teachers, designers, domain experts — none of these require you to be especially mathy. If math isn’t your strength, pick a path where it isn’t required. The field is vastly wider than the mathematical core.

Myth “You need a computer science degree.”

Reality For a subset of roles, yes. For the majority of AI-adjacent careers, no. Many of the most successful people in applied AI come from business, social sciences, design, healthcare, and the humanities. The degree matters less than your ability to learn and apply.

Myth “It’s too late to start — everyone’s been doing this forever.”

Reality Nobody has been doing modern AI for long. The LLM era is about three years old. The field is young. Most experts are people who got interested between 2022 and 2024 and put in work. If you start now, in 2026, you’re not late. You’re early for everything that comes next.

Myth “AI will automate its own job market before I get in.”

Reality It’s reshaping some roles, expanding others, and creating new ones. The AI industry is hiring aggressively — especially for applied, deployment, and governance roles. The concern that AI will eat all AI jobs is worth taking seriously, but the current trajectory is growth, not collapse.

Myth “Small towns and regular cities don’t have AI jobs.”

Reality Less true every month. Remote work and distributed teams mean AI roles exist far from Silicon Valley. More importantly, the Path 3 (domain application) careers are everywhere — hospitals, schools, local government, small businesses, insurance offices, law firms. Your hometown has more AI-adjacent jobs than you think.

§07 Career-path terms worth knowing

A career-specific glossary.

Builder / implementer / practitioner / steward

The four broad AI career paths — described in §02 above. Useful shorthand for thinking about where you fit.

ML Engineer vs. AI Engineer

ML engineers build and train models (research-adjacent). AI engineers typically integrate existing models into products and systems (application-focused). The distinction isn’t perfect, but it matters when reading job descriptions.

Prompt engineer

A role focused on getting the best output from existing AI models through careful prompting. Emerged rapidly, may be fading as models get better — but the underlying skill (clear thinking about how to communicate with AI) will stay valuable.

Solutions engineer / customer engineer

A role that bridges sales, engineering, and customer implementation. Underrated path into AI careers for people who like variety and client work.

Research scientist

A role at AI labs (Anthropic, Google DeepMind, OpenAI, Meta AI, etc.) focused on advancing the frontier of what AI can do. Typically requires a PhD or equivalent research experience.

Applied AI / Applied ML

Work that uses existing AI techniques to solve real business problems. Distinguished from “research” or “frontier” work. The larger and more accessible category.

AI PM (Product Manager)

Decides what AI products and features get built. One of the most in-demand non-engineering roles in tech.

AI Ops / MLOps

The operational side of running AI systems in production — deployment, monitoring, scaling, debugging. A growing field with strong job security.

Red teaming

Deliberately trying to break AI systems to find vulnerabilities. A growing specialty in AI safety and security.

AI governance

The broader field of ensuring AI is developed and used responsibly. Includes policy, ethics, compliance, and auditing.

Domain expert

Someone with deep knowledge in a specific field (healthcare, law, education, finance, etc.) who understands how AI applies to that field. Increasingly valuable.

Adjacent careers

Careers that aren’t “in AI” but are heavily shaped by it. The largest category, and the one most overlooked by students focused on tech-sector jobs.

Side project / portfolio

Public, demonstrable work you’ve built outside your formal coursework or job. Increasingly required for competitive AI roles.

Certifications

Credentials from vendors (AWS, Google, Microsoft) or industry bodies (ISACA, ISC2) that demonstrate specific skills. Valuable for career switchers and for deployer/practitioner/steward paths.

§08 What to carry forward

Seven things to remember as you make career choices.

“Careers in AI” is a wide category, not a narrow one. The builders get the headlines. The practitioners make up most of the field.
Your domain matters as much as your technical skill. Being deeply good at one field (healthcare, education, law, finance, trades) plus being AI-fluent is a rarer and more valuable combination than being a generalist AI person.
Pick a major you love in a field you want to spend time thinking about. The specific major matters less than your ability to learn, build, and communicate.
Build something public. A project, a blog, a case study, a presentation. Something a hiring manager can see. “I’m interested in AI” is invisible. Evidence is not.
Say yes to the AI work at your first job. Many serious AI careers started with someone raising their hand.
Talk to real people doing the jobs you’re considering. Five 20-minute conversations beat five months of scrolling. Most professionals will say yes to a thoughtful student.
It’s not too late. Modern AI is three years old. You are on time for the field that matters next.

The best career advice I’ve ever heard applies perfectly here: pick something you can be interested in for ten years, become genuinely good at it, and stay curious. The specific job titles will change. The habits won’t.

The next card in the series gets into the most transferable skill you can build in this era: AI literacy. Not just knowing what AI is, but developing the habits — prompt thinking, output evaluation, knowing when not to use it — that make you genuinely competent with these tools. If you’ve built the foundation from the first four cards, this is the one that turns understanding into practice.

Next in this series

Building AI Literacy

What AI literacy actually means in practice — prompt thinking, evaluating outputs critically, knowing when not to use AI, and how to build these skills deliberately before you need them on the job.