What the New AI Job Titles Actually Mean — and Which Ones Are Technical

Half of my job-alert feed now has "AI" in the title. That is not a figure of speech — of the last 81 roles LinkedIn sent me, 45 carried AI, ML or generative in the title. The more telling number is this one: those 45 postings used 40 distinct titles. Almost no two employers call the same job the same thing.

Sit with that for a second. The market has not converged on a vocabulary, which makes the title the least reliable line in any AI job ad. Candidates apply into the wrong rooms; hiring managers run interview loops calibrated for a different job than the one they are filling. I have spent fifteen years in product and technology hiring conversations, mostly in regulated fintech, and I have never seen the labels drift this far from the work.

So here is the map I wish someone had handed me: each of the emerging titles, what it actually is, whether it is genuinely new or an old role wearing an adjective, whether it is a technical seat — and, since ads are cagey about money, which long-established role it sits beside in seniority and pay. Deliberately, no salary figures. Figures date within a quarter; the mapping holds.

The test that works

For every title below I apply the same test: delete the word "AI", read the responsibilities, and ask which role from the last twenty years is left standing. Usually there is a clear answer, and that answer — not the title — tells you the seniority, the interview loop and the money. The few cases where nothing is left standing are the genuinely new jobs, and those are worth knowing by name.

Forward Deployed Engineer — technical, and more junior than it sounds

This is the title people ask me about most, usually assuming it is some new species of senior hybrid. It is not. A Forward Deployed Engineer is a software engineer who embeds in a customer's organisation: sits with the client, does the discovery, and personally writes the code that makes the product work inside that client's messy reality. Palantir ran this model for years; the current AI wave has generalised it — postings for it now run into the thousands across hundreds of companies, with the bulk of them in the US.

Make no mistake about the bar. Across the postings I have read, Python is near-ubiquitous and customer-facing work is the single most-listed responsibility. This is production coding plus consulting in one seat, often with real travel — regular travel is the norm rather than the exception.

And here is what the glamour of the title hides: it is mostly a mid-level job. Most postings ask for a mid-career amount of experience, and very few ask for a decade or more. In known-role terms, an FDE sits where a mid-level full-stack engineer sits, and is paid accordingly — not like the exotic hybrid the title suggests. If you are a strong engineer a few years in who likes clients, it may be the best door in the industry right now. If you are a fifteen-year technology leader, it is a step down wearing a novelty hat.

Technical: yes — hands-on production code, non-negotiable. New or rebadge: genuinely new at scale. Sits like: a mid-level full-stack engineer, embedded at the customer.

Forward Deployed Product — the same shape, without the coding

This is the most useful distinction in the whole landscape, and almost nobody makes it. The forward-deployed model — embed with the customer, own the deployment, feed what you learn back into the core product — now also ships with a product leader in the seat instead of an engineer. The titles read Forward Deployed Product Manager, or Director of Product Management, Forward Deployed.

Read the actual requirements and the difference from FDE is stark. Scale AI's London posting for the director version asks for ten-plus years of product leadership, including four-plus years managing PM teams, and states the technical bar as "high technical IQ: comfort with APIs, data pipelines, SQL" — with coding explicitly not required. Cresta's UK-remote version asks for "strong technical understanding, systems thinking and exposure to building Agents (not necessarily conversational, personal projects count!)". These companies want someone who can sit across the table from a Fortune 100 CIO's team, gather requirements, and make correct technical judgements about what the deployment needs — while directing the build rather than typing it.

In known-role terms this is a senior product director or delivery lead who lives with clients, with seniority and pay to match — a full band above where FDE sits. One honest caveat: the market is thin. Exact-phrase searches on the big UK boards return almost nothing, and I could find fewer than a dozen live postings across the UK and Europe. It is a set-an-alert title, not a volume channel. But it is growing out of the fastest-scaling companies in the industry, and its shape — technical judgement without the keyboard — is where I would bet several of these titles eventually converge.

Technical: no code — but real technical fluency, tested hard. New or rebadge: genuinely new, and barely mapped. Sits like: a senior product director who lives with clients.

Head of AI — one title, three different jobs

Head of AI has quietly become the mainstream senior AI title — UK postings nearly tripled year on year in the series I track. It has also become three different jobs sharing one name, and applying to the wrong one is the single most common mistake I see people make:

• The technical Head of AI, mostly at AI-native companies: applied research and platform. The ads say it plainly — an engineering or research background and experience building or deploying large language models in production. This is a head-of-engineering seat; it sits like a CTO-track engineering leader, and no amount of adjacent experience substitutes for having shipped models.
• The governance Head of AI, in regulated industries: model risk, auditability, an AI posture a regulator will accept. It needs a working technical understanding — you cannot govern what you cannot read — but not a research record. It sits like a senior risk-and-technology leader.
• The adoption Head of AI, everywhere else: getting a company to actually use this stuff. Tooling, vendor calls, training, judgement about what to trust. The hiring criterion is commercial judgement, not research pedigree. It sits like a transformation or delivery director with a strong technical spine.

The function itself is usually small — commonly one to ten people. The tell for which archetype you are reading is always in the requirements paragraph, never in the title. And the volume boom carries a warning worth stating dryly: as the title gets easier to acquire, the new volume is arriving at the bottom of the band rather than the top, and the same three words on a CV mean a little less each quarter. Which of the three jobs you would actually be doing matters more than ever.

Technical: one archetype of three; the other two need fluency, not code. New or rebadge: the name is new; two of the three jobs are not. Sits like: engineering leader / risk-and-technology leader / transformation director — read the body to find out which.

VP of AI, AI Director — Head of AI in a different org chart

Same job, one rung up or inside a US-headquartered structure that spends the VP word more freely. Everything above about the three archetypes applies unchanged. If the requirements read like applied research it is the technical seat; if they read like change management it is not.

Technical: varies with the archetype underneath. New or rebadge: a rebadge of Head of AI. Sits like: the same seats, retitled.

Chief AI Officer — a real seat, barely a job market

Chief AI Officer is the title the press writes about most and the job boards carry least. Across the AI-leadership hiring I have watched this year I have not seen one come open. The only one to reach my own feed in months was posted by an executive search firm, not an employer. Chief-level AI postings have not grown the way senior AI-leadership postings below them have.

The explanation is simple: CAIO is a seat people are appointed or headhunted into, almost always after visibly owning an AI mandate somewhere for a few years. In known-role terms it is a CTO-peer executive seat, and the technical question splits the same three ways as Head of AI, one level up. If the plan is to become one, the route runs through a Head of AI or VP of AI seat and two or three years of owning the mandate — not through an application form.

Technical: varies — the same three archetypes, at board level. New or rebadge: new, and mostly appointed rather than advertised. Sits like: a CTO-peer executive.

AI Product Manager, Head of Product (AI) — the honest rebadge

These are the easiest to decode, because they are exactly what they say: product management where the product happens to be AI. The job has not changed — roadmap, trade-offs, stakeholders, shipping. The subject matter has: evaluation methodology instead of pixel-perfect specs, model build-versus-buy calls, and the strange new discipline of shipping products that are non-deterministic by design.

Not a coding seat — but the fluency bar is real and rising. A good AI product person can reason about why an eval suite is lying, what a retrieval pipeline can and cannot be blamed for, and when a model swap is a product decision rather than an engineering one. An AI Product Manager sits where a senior product manager sits; a Head of Product for an AI product sits where a product director sits. Watch the seniority word carefully: the same "AI product" phrasing appears on ads a full band apart, and the adjective does not add a band.

Technical: no code; genuine fluency required. New or rebadge: a rebadge, now well established. Sits like: a senior PM or a product director — the seniority word decides, not the AI word.

AI Transformation, Enablement, Adoption — the biggest category, and the least technical

By volume this is the largest emerging category in every dataset I have looked at — it is the majority of what I see advertised under an AI-leadership banner. Head of AI Transformation, Group Head of AI Enablement, Director of AI Adoption: strip the adjective and what is left standing is change leadership. Getting a large organisation to genuinely use AI — tools rolled out, people trained, workflows redesigned, adoption measured — is the job digital-transformation leaders have been doing for two decades with a different noun on the door.

That is not a criticism. It is where an enormous amount of the real value sits, because the binding constraint on AI in most companies is not the models — it is the organisation. But be clear-eyed about the seat: it is not technical and does not pretend to be, and it sits where a transformation programme director sits, typically a notch below where the same company's Head of AI sits. One caution from reading a lot of these ads: some carry a genuine product mandate and some are programme management wearing a product title. The body tells you; the title never does.

Technical: no. New or rebadge: new as a name, a rebadge of transformation leadership in substance. Sits like: a transformation programme director.

Field CTO — the C is doing a lot of work

A Field CTO is a vendor's most senior technical voice in front of customers' executives. The tell is in one vendor's own filing: Cloudera lists its Field CTO openings under Sales Engineering. This is the elevated form of the principal solutions architect or pre-sales lead — deeply technical in conversation, rarely building anything, measured on revenue influence. It is real and growing (I count over a dozen live UK listings across eleven employers) and it is genuinely senior. But it is not the executive seat the name implies, and if your background is client-side technology leadership rather than vendor-side pre-sales, expect the lineage question in the first interview: every posting I have read wants prior customer-facing vendor scar tissue.

Technical: yes, in depth of conversation — not in code shipped. New or rebadge: a rebadge of principal solutions architect, elevated. Sits like: the most senior pre-sales engineer in the company.

AI Solutions Director, Head of AI Advisory — the consultancy chapter

Consultancy-side AI practice leadership: sell the work, scope it, staff it, land it. This is the practice-lead role that professional services firms have always run, pointed at AI engagements. Partly technical in the pre-sales sense, with the same lineage expectations as Field CTO one seat over. Good work if selling is in you; a misfit if what you actually want is to build.

Technical: partly — enough to scope honestly. New or rebadge: a rebadge of consulting practice leadership. Sits like: a practice director at a consultancy.

AI Governance Lead — the quiet one

Policy, model risk, auditability, EU AI Act and regulator exposure. New as a named title, but directly adjacent to the model-risk management function regulated firms have run for years. Not a coding seat; it needs enough technical depth to read a model card critically and push back on an engineering team's assurances.

Technical: fluency, not code. New or rebadge: new name, near-adjacent to model-risk management. Sits like: a senior compliance or model-risk lead.

The title that does not exist

While researching this piece I went looking for "Head of Generative Content", a title I had seen suggested as an emerging one. It returned zero postings. Not few — zero, across LinkedIn, the major UK boards and general web search. The nearest real jobs are agency-side creative leadership roles asking for a decade of creative agency and content-production experience plus hands-on generative craft — a different career entirely.

I include this partly as a public service and partly as a caution: in a market inventing titles this fast, some of the titles in circulation are not attached to any actual jobs. Before repositioning a career toward a title, check that the postings exist.

The cheat sheet

What "technical" actually means now

The word doing the most damage in all of this is "technical", because job ads use it to mean two different things and rarely say which. Sometimes it means writes production code — the Forward Deployed Engineer, the technical Head of AI. Sometimes it means can be trusted with technical judgement: reads code and systems well enough to direct them, argue with them, and catch them lying, without being the person at the keyboard. Scale AI's "high technical IQ, coding not required" is the most honest phrasing of the second kind I have seen in a live posting.

The industry is slowly noticing that it needs both kinds, and that they are different people. The first kind is easier to interview for, so ads over-index on it even when the seat plainly needs the second. My advice, having watched this from both sides of the table: decide which kind you are, say it plainly, and match yourself on the body of the ad rather than the label on it. The titles will keep churning for a few more years. The jobs underneath them are older, steadier, and easier to navigate than they look.

This essay also lives at Appaya, the AI product studio I run.