AI Careers v1 · 15 roles

The new careers AI is opening up

Fifteen emerging roles where AI is the work, the tool, or the daily colleague. Each one chosen against four hard criteria — not invented from a buzzword list.

How these 15 made the list

Every role on this page passed our AI Career Inclusion Framework: it has measurable hiring demand (job-posting threshold met), at least two independent citations (WEF Future of Jobs, LinkedIn Economic Graph, Stanford HAI AI Index, ESCO, Anthropic Economic Index, or a national labour statistics agency), maps to the six career-capability constructs the PositionMeAI platform is built on, and includes an honest statement of what AI augments vs what stays human.

If a role isn't here, it's because it failed one of those tests — not because we missed it. "AI Whisperer," "Prompt Wizard," and similar fabrications are intentionally excluded. Re-reviewed annually each Q1; next review June 2027.

WEF 2026 listed

AI Product Manager

The translator between AI capability and human need.

Owns the shape, success metrics, and roadmap of an AI-powered product. Less code-writing, more judgement: which problems are worth using AI on, which user signals matter, where the model's limits become product risks. Sits between research/ML teams and customers.

$6,500–$14,000 / mo ~4,200 active postings
AI impactAI accelerates prototyping and user-research synthesis. AI cannot decide what to build or judge product-market fit — that requires direct contact with real users and stakeholder politics.
Citations & GSAS mapping
  • Dominant constructs: Goal-Directed Behaviour, Career Self-Management, Adaptability
  • Stretched by AI: Critical Reflective Skills
  • WEF Future of Jobs Report 2026
  • LinkedIn Economic Graph: Jobs on the Rise 2025
  • Stanford HAI AI Index Report 2026, Ch. 4
  • Entry routes: Bachelor's + 3-5y adjacent PM/eng; MBA with ML; move from adjacent-domain PM
WEF 2026 listed

ML Ops Engineer

Keeps the models running in production — the unsung backbone.

Builds the pipelines that get ML models from notebook to live system and keep them healthy. Versioning, deployment, monitoring, retraining triggers, cost optimization, latency budgets. Where most AI projects quietly fail without this role.

$7,000–$16,000 / mo ~52,000 active postings
AI impactAI assists with code generation and incident summarisation. AI cannot judge whether a model degradation is acceptable or warrants rollback — that decision still requires human context.
Citations & GSAS mapping
  • Dominant: Lifelong Learning, Goal-Directed Behaviour
  • Stretched by AI: Adaptability
  • WEF Future of Jobs Report 2026
  • LinkedIn Economic Graph 2025: AI Engineering Operations
  • US BLS Occupational Outlook 2026 (specialty: ML infrastructure)
  • Entry routes: CS/SWE bachelor's + 2-3y backend/DevOps; ML eng cert + prod experience; SRE transition
WEF 2026 + LinkedIn #1 growth

Generative AI Engineer

Builds applications on top of large language and image models.

Specifically applies foundation models (GPT-class, Claude-class, Gemini-class, image/audio/video) to real product problems. Designs retrieval, agents, evaluation harnesses, guardrails. Different from ML Engineer (who often trains models from scratch) and Prompt Engineer (whose scope has largely been absorbed into this role).

$7,500–$18,000 / mo ~8,400 postings · +74% YoY
AI impactAI is the medium and the colleague — this role lives inside AI capability. The non-automatable parts are judgement about model selection, evaluation design, and failure-mode anticipation.
Citations & GSAS mapping
  • Dominant: Lifelong Learning, Critical Reflective Skills, Adaptability
  • Stretched by AI: Lifelong Learning
  • LinkedIn Economic Graph 2025: 74% YoY growth
  • WEF Future of Jobs Report 2026, p. 47
  • Anthropic Economic Index 2025
  • Entry routes: CS/ML bachelor's + LLM-applied portfolio; SWE + foundation-model training; research scientist transition
WEF 2026 listed

AI Ethics Specialist

Decides what AI should and shouldn't do, with teeth.

Develops, maintains and enforces an organisation's AI use policies. Reviews proposed AI applications for harm potential, bias, accessibility, regulatory exposure. Sits in legal, risk, or product depending on company — increasingly a standalone function as the EU AI Act enforcement matures.

$6,500–$13,000 / mo ~1,200 active postings
AI impactGenerative AI scales the volume of content and decisions needing review, making the role bigger not smaller. AI can flag candidates for review but cannot decide policy edge cases — that judgement remains human.
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Social Capability, Career Self-Management
  • Stretched by AI: Critical Reflective Skills
  • WEF Future of Jobs Report 2026
  • Stanford HAI AI Index Report 2026, Ch. 6
  • ESCO: Ethics Officer (AI specialisation)
  • Entry routes: Law + tech; philosophy/ethics + eng exposure; product/policy + IAPP AIGP cert
WEF 2026 listed

AI Trust & Safety Specialist

Catches what shouldn't make it out of an AI system.

The operational counterpart to AI Ethics. Builds and runs content review, red-teaming, abuse-pattern detection, and incident response for AI-generated output. Often inside the platform/product team. Hires exploded post-2023 due to generative AI scaling content volume by orders of magnitude.

$5,500–$12,000 / mo ~2,800 active postings
AI impactAI dramatically scales the volume of content/output needing review. Adversaries also use AI to generate evasive content. Human judgement remains essential for ambiguous and culturally-specific cases.
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Adaptability, Social Capability
  • Stretched by AI: Adaptability
  • LinkedIn Economic Graph 2025: Trust & Safety sub-domain
  • Stanford HAI AI Index Report 2026, Ch. 5
  • Anthropic Economic Index 2025
  • Entry routes: T&S background + AI training; content-moderation + policy work; security research
Established · LinkedIn 2025

AI Solutions Architect

Designs how AI gets stitched into existing enterprise systems.

Senior technical role. Designs the architecture for integrating AI capabilities into existing business systems — what model, hosted where, how authenticated, how it handles cost spikes, how it falls back when AI is unavailable. Usually customer-facing, bridges engineering and sales.

$8,000–$18,000 / mo ~6,300 active postings
AI impactAI shifts what's possible in system design weekly. AI can generate reference architectures but cannot understand the political and legacy constraints of a specific organisation — that's the work.
Citations & GSAS mapping
  • Dominant: Goal-Directed Behaviour, Social Capability, Adaptability
  • Stretched by AI: Lifelong Learning
  • LinkedIn Economic Graph 2025
  • US BLS: Computer Network Architect (AI specialisation)
  • Entry routes: 10+y SW architecture + AI/ML deep-dive; cloud architect + AI; solutions eng progression
Normalising · LinkedIn 2024

Prompt Engineer / AI Interaction Designer

Designs how humans and AI models actually talk to each other.

Originally exploded in 2023, the role has normalised. Pure prompt-writing has been largely absorbed into Generative AI Engineer scope. The surviving discipline is broader: designing the full interaction pattern between human and AI — chains, agents, error recovery, when to defer to humans, when to push back.

$5,500–$11,000 / mo ~1,100 postings (declining)
AI impactAI is the entire job. The role exists because AI is unreliable in specific reproducible ways. As models improve, the role shrinks in some areas (raw prompt tuning) and grows in others (agent design, evaluation).
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Adaptability
  • Stretched by AI: Critical Reflective Skills
  • LinkedIn Economic Graph 2025: AI Interaction Design
  • Anthropic Economic Index 2025
  • Entry routes: Strong writing + AI fluency; UX writing + AI tooling; research methods/linguistics
Established · LinkedIn 2024

Conversational AI Designer

Designs voice and chat interfaces that don't feel robotic.

Designs voice and chat experiences where AI is the interface. Covers chatbots, voice agents, customer-service automation, accessibility interfaces. Less about model selection (that's engineering) and more about flow design, escalation logic, tone, and fail-state experience.

$5,000–$10,000 / mo ~1,800 active postings
AI impactAI made conversation interfaces actually work, expanding the role. AI cannot judge whether a tone feels patronising or supportive to a specific user audience — that's still human craft.
Citations & GSAS mapping
  • Dominant: Social Capability, Critical Reflective Skills, Adaptability
  • Stretched by AI: Social Capability
  • LinkedIn Economic Graph 2025
  • ESCO: User Experience Designer (conversational specialty)
  • Entry routes: UX/service design + AI fluency; linguistics/comms + UX training; CX leadership transition
Established · multimodal-driven growth

Computer Vision Engineer

Teaches machines to see — newly central in the multimodal era.

Long-established role that's resurged with multimodal AI. Builds systems that interpret images, video, satellite, medical imaging, sensor data. Diagnostic imaging, autonomous driving, agriculture, retail, defence, accessibility. Pre-dates the gen-AI wave but is now revitalised by foundation vision models.

$7,000–$16,000 / mo ~9,500 active postings
AI impactFoundation vision models (CLIP, Gemini, GPT-4V class) raised the floor for what's possible, shifting the work from training-from-scratch to careful fine-tuning, evaluation, and domain adaptation.
Citations & GSAS mapping
  • Dominant: Lifelong Learning, Critical Reflective Skills
  • Stretched by AI: Lifelong Learning
  • US BLS Occupational Outlook 2026
  • Stanford HAI AI Index Report 2026, Ch. 2
  • ESCO: Software Engineer (computer vision specialisation)
  • Entry routes: CS/ML bachelor's-master's + CV; image processing + DL training; research lab transition
WEF 2026 listed

AI Research Scientist

Pushes the boundary of what AI can actually do.

Industrial or academic research role. Builds new methods, evaluates capabilities, publishes. Increasingly hosted in private labs (Anthropic, DeepMind, OpenAI, Meta AI, smaller specialised labs) as compute scale has moved beyond most universities. PhD typical but not universal — exceptional portfolios can replace it.

$10,000–$30,000+ / mo ~1,900 active postings
AI impactAI accelerates literature review, idea generation, and proof-of-concept code. AI cannot identify which research questions actually matter or judge whether a finding is novel — that requires deep field immersion.
Citations & GSAS mapping
  • Dominant: Lifelong Learning, Critical Reflective Skills, Career Self-Management
  • Stretched by AI: Critical Reflective Skills
  • Stanford HAI AI Index Report 2026, Ch. 1, 4
  • WEF Future of Jobs Report 2026
  • US BLS: Computer & Information Research Scientists
  • Entry routes: PhD in ML/AI/CS; exceptional portfolio + bachelor's-master's; postdoc-industry transition
WEF 2026 listed · emerging

Synthetic Data Engineer

Generates training data that's safer, more diverse, or otherwise impossible to collect.

Builds systems that generate synthetic training data when real data is unavailable, privacy-restricted, or biased. Critical in healthcare (privacy), autonomous driving (rare events), fraud detection (rare positives), and now in foundation-model post-training. Newer specialisation, accelerating.

$7,000–$15,000 / mo ~600 active postings
AI impactAI generates the data. The job is judging whether the synthetic data actually represents the real distribution — a deeply non-automatable judgement.
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Lifelong Learning
  • Stretched by AI: Critical Reflective Skills
  • WEF Future of Jobs Report 2026 — emerging roles
  • Stanford HAI AI Index Report 2026, Ch. 3
  • Entry routes: ML eng + statistical depth; data eng + generative model fluency; research scientist transition
WEF 2026 listed

AI Policy Analyst

Bridges technical reality and regulatory language.

Works in government, think tanks, industry-bodies, or large companies' policy teams. Reads what AI labs publish, translates it into regulatory language and risk assessments. Demand exploded post-EU AI Act 2024 and continues to grow as more jurisdictions (UK, US states, ASEAN, AU/NZ) develop AI law.

$5,500–$13,000 / mo ~1,400 active postings
AI impactAI accelerates literature review and comparative-policy synthesis. AI cannot judge stakeholder politics, regulatory enforceability, or read between the lines of what regulators actually want.
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Social Capability, Career Self-Management
  • Stretched by AI: Critical Reflective Skills
  • Stanford HAI AI Index Report 2026, Ch. 6
  • WEF Future of Jobs Report 2026
  • OECD AI Observatory 2025
  • Entry routes: Public policy/law + tech literacy; technical role + policy training; PhD in policy/econ/law
WEF 2026 · EU AI Act-driven

AI Auditor / Algorithmic Auditor

Independent verification that an AI system does what it claims.

Emerging compliance role. Audits AI systems against regulatory frameworks (EU AI Act, sector-specific rules), against company policies, or for third-party assurance. Demand accelerated sharply when EU AI Act enforcement provisions kicked in (Aug 2026). Often inside large audit firms or specialist consultancies.

$6,000–$14,000 / mo ~850 active postings
AI impactAI can generate test cases and probe model behaviour at scale. AI cannot judge whether a system meets the spirit of a regulation — that requires legal interpretation and stakeholder context.
Citations & GSAS mapping
  • Dominant: Critical Reflective Skills, Goal-Directed Behaviour, Career Self-Management
  • Stretched by AI: Critical Reflective Skills
  • WEF Future of Jobs Report 2026
  • Stanford HAI AI Index Report 2026, Ch. 6
  • ISO/IEC 42001:2023 AI Management Systems Standard
  • Entry routes: Audit firm + AI/ML depth; compliance + technical training; infosec progression
Established · bifurcating

AI Training Data Specialist

Curates, labels and quality-checks the data models learn from.

Often globally distributed remote work. Specialises in annotating training data for specific domains — medical, legal, multilingual, niche language pairs, expert-judgement RLHF. The work split between contract platforms (Scale, Surge, Outlier, Labelbox) and in-house teams. Mid-skill entry point into AI work; transparent salary varies widely by country.

$800–$5,500 / mo (bimodal) ~12,000 active postings
AI impactAI labels much of what doesn't need expert judgement. Human labellers increasingly focus on adversarial, edge-case, expert-domain, or culturally-specific work. Compensation is bimodal — generalist labelling has compressed, specialist labelling has risen.
Citations & GSAS mapping
  • Dominant: Goal-Directed Behaviour, Critical Reflective Skills
  • Stretched by AI: Critical Reflective Skills
  • LinkedIn Economic Graph 2025
  • Anthropic Economic Index 2025: RLHF and supervised fine-tuning labour
  • Entry routes: No degree for entry-level; expert specialisations need credentials in that field; QA/content-moderation transition
WEF 2026 listed · evolving

AI-Augmented Healthcare Specialist

Clinical work where AI is now a daily tool, not a future threat.

Existing clinical roles (radiologist, pathologist, dermatologist, general practitioner) where AI tools are now part of standard practice. Not a separate job title in most healthcare systems — more an evolution of existing roles. Listed here because the GSAS construct demands and entry route are now meaningfully different.

$8,000–$25,000+ / mo Existing clinical roles
AI impactAI tools assist with image diagnosis, triage, and pattern detection. AI cannot take clinical responsibility, judge ambiguous cases, or perform the relational and physical-examination dimensions of care.
Citations & GSAS mapping
  • Dominant: Lifelong Learning, Critical Reflective Skills, Career Self-Management
  • Stretched by AI: Critical Reflective Skills
  • WEF Future of Jobs Report 2026
  • Stanford HAI AI Index Report 2026, Ch. 4, 5
  • WHO Global Strategy on Digital Health
  • Entry routes: Standard medical/clinical pathway + AI tooling fluency in residency/CPD; established clinicians upskilling via structured AI-in-healthcare programmes

Want to see if any of these fit you?

Take the free Career Report — it maps your profile against the six career-capability constructs and shows which directions actually align with your evidence. No assumptions, no AI guesses about you.

Take the Career Report → I'm still in school