AI Jobs in 2027: 9 New Roles, 9 Changed Roles, and What They Pay in the US, UK, Germany and Israel
The new AI roles of 2027, the established roles that changed, and a salary comparison in US dollars across four AI hubs. An ML or AI engineer earns a median of $283,000 in the US, $184,000 in Israel, $137,000 in the UK and $100,000 in Germany.
2027 is the year AI roles leave the engineering department and spread across the whole organization. Two years ago an "AI job" almost always meant an engineer or a researcher. Today there are Chief AI Officers on bank leadership teams, managers responsible for agents, and compliance people who became AI governance leads. At the same time, established roles such as data scientist, QA tester and customer service agent have changed what they do, and sometimes their name. Below are the 9 new roles, the 9 roles that changed, and a salary comparison in US dollars across four AI hubs: the United States, the United Kingdom, Germany and Israel.
- According to IBM, 76% of organizations have appointed a Chief AI Officer, up from 26% a year earlier.
- AI engineer is the fastest-growing job in the US according to LinkedIn, with AI consultant in second place.
- A machine learning or AI engineer earns a median total compensation of $283,000 in the US, $184,000 in Israel, $137,000 in the UK and $100,000 in Germany.
- The "AI premium" over a regular software engineer is largest in the US (about 43%) and smallest in Germany (about 5%).
- Prompt engineer is barely a standalone job anymore: the skill was absorbed into other roles.
Why 2027 is the turning point for AI roles
Three data points explain the shift. The first is the pace of executive appointments: IBM's CEO study, run from February to April 2026 among 2,000 CEOs in 33 countries, found that 76% of organizations have appointed a Chief AI Officer. The second is engineering demand: according to LinkedIn's Jobs on the Rise 2026, AI engineer is the fastest-growing job in the United States and AI consultants are second. The third is the long-term forecast: the World Economic Forum's Future of Jobs Report lists AI and machine learning specialists among the three fastest-growing roles through 2030.
AI salaries in 2027: US vs. UK vs. Germany vs. Israel (USD)
To compare markets fairly we used one source for all four countries: median total compensation (base salary plus annualized stock plus bonus) reported to Levels.fyi, updated October 24 to 25, 2026. We converted pounds, euros and shekels to dollars at the European Central Bank reference rates of October 23, 2026 (1 USD = 0.748 GBP, 0.893 EUR, 3.12 ILS), and rounded to the nearest thousand.
| Role (median total comp) | US | UK | Germany | Israel | 4-hub average |
|---|---|---|---|---|---|
| ML / AI Engineer | $283,000 | $137,000 | $100,000 | $184,000 | $176,000 |
| Product Manager | $231,000 | $140,000 | $106,000 | $150,000 | $157,000 |
| Data Scientist | $180,000 | $112,000 | $87,000 | $153,000 | $133,000 |
| Data Engineer | $159,000 | $106,000 | $88,000 | $124,000 | $119,000 |
| Software Engineer (baseline) | $198,000 | $120,000 | $95,000 | $145,000 | $140,000 |
Sources: Levels.fyi pages for ML engineers in the US, UK, Germany and Israel, and US pages for product managers, data scientists, data engineers and software engineers (other countries via the same pages' location filters).
How to read the table:
- The AI premium differs by market. An ML or AI engineer earns about 43% more than a regular software engineer in the US, about 27% more in Israel, about 14% more in the UK and only about 5% more in Germany. The US premium reflects how much equity big US tech companies pay for AI talent.
- Israel is the second-highest market for AI engineers in this comparison, ahead of the UK and Germany, in line with its density of AI startups and global R&D centers.
- These are tech-company numbers. Levels.fyi data leans toward technology companies that pay in stock. Salaries at banks, government and traditional companies are usually lower, and base-salary-only surveys show smaller figures.
- Sample sizes vary. The US ML engineer median is based on more than 1,000 reports, Israel's on 26. Treat the smaller markets as directional.
For the new non-engineering roles, such as Chief AI Officer, AI governance lead or agent specialist, there is no reliable public benchmark that compares countries. Rather than invent numbers, each role below says which pay level it is closest to.
9 new AI roles
Each role follows the same structure: what they do, requirements, where you find it, pay, who moves into it, and expected demand.
1. Chief AI Officer (CAIO)
The role that best symbolizes the shift. In IBM's spring 2026 CEO study, 76% of surveyed organizations had appointed a Chief AI Officer, up from 26% a year earlier, one of the fastest expansions ever seen in the C-suite.
| What they do | Own the AI strategy: where to invest, which systems to approve, how to measure business value, and how to govern risk. Usually reports to the CEO. |
| Requirements | Senior leadership experience, deep technology understanding (coding is optional), business fluency, regulatory awareness and the ability to lead organizational change. |
| Where you find it | Banks, insurers, large enterprises and government. |
| Pay level | At senior VP or C-suite level, usually with bonus and equity. There is no reliable public cross-country benchmark for this title yet, so we do not quote a number. |
| Who moves into it | CTOs, chief data officers, digital transformation leaders and senior product leaders. |
| Demand in 2027 | Very high in large organizations. In smaller companies it is usually an extra hat for the CEO or CTO. |
2. AI Engineer
The fastest-growing job in the US according to LinkedIn's Jobs on the Rise 2026. An AI engineer does not train models from scratch; they build products and workflows on top of existing language models.
| What they do | Connect language models to company systems, build retrieval (RAG) and agents, write evaluations, and manage cost, latency and security. |
| Requirements | Python, software engineering, model APIs, RAG, frameworks such as LangChain and PyTorch, cloud and Docker. |
| Pay (median total comp, USD) | US $283,000, UK $137,000, Germany $100,000, Israel $184,000. See the full table above (Levels.fyi). |
| Who moves into it | Backend and full-stack developers, ML engineers and data scientists. |
| Demand in 2027 | Very high in every market we compared. |
3. Forward Deployed Engineer (FDE)
The role that the word "pilot" created. According to Lightcast, US job postings went from about 200 in 2024 to about 1,200 in 2025 and more than 5,200 in the first seven months of 2026.
| What they do | Sit with the customer, inside their real systems and data, and turn an AI product into a workflow that runs in production. Half engineer, half consultant. |
| Requirements | Software and AI skills, business process understanding, excellent customer communication and comfort with ambiguity. |
| Pay level | Usually at the level of a senior AI engineer, sometimes with a premium for customer-facing work. No separate public benchmark yet. |
| Who moves into it | Solutions engineers, AI engineers and technical consultants. |
| Demand in 2027 | High and rising. People who can both code and talk to a customer are the bottleneck. |
4. AI Product Manager
A product manager whose product is not deterministic. Instead of specifying buttons, they define what a good answer is, how to measure it and what happens when the model is wrong.
| What they do | Define the business problem, decide build vs. buy, design the AI user experience, and own quality and cost metrics. |
| Requirements | Product management experience, understanding of model capabilities and limits, evaluation thinking and privacy awareness. |
| Pay (median total comp, USD) | No separate "AI PM" benchmark. Product managers overall: US $231,000, UK $140,000, Germany $106,000, Israel $150,000 (Levels.fyi). AI specialization typically adds a premium on top. |
| Who moves into it | Experienced product managers, analysts and UX professionals. |
| Demand in 2027 | High. Almost every product is adding AI, and someone has to own its quality. |
5. AI Governance and Risk Lead
According to the IAPP AI Governance Profession Report, 77% of organizations are already working on AI governance. Someone has to own it day to day.
| What they do | Run the AI system inventory, risk assessments, policies and incident procedures, and ensure compliance with privacy law and the EU AI Act. |
| Requirements | Background in privacy, compliance, risk or law, basic technical understanding, and familiarity with ISO/IEC 42001 and the EU AI Act. |
| Pay level | Similar to senior compliance and risk roles, higher in regulated industries. No reliable cross-country public benchmark yet. |
| Who moves into it | Privacy officers, compliance and risk professionals, lawyers and security staff. |
| Demand in 2027 | Rising, driven largely by EU AI Act transparency obligations that took effect in August 2026. |
6. AI Agent Specialist (Agent Ops)
In Microsoft's Work Trend Index, 32% of leaders said they plan to hire AI agent specialists in the next year, and 28% are considering managers for blended teams of people and agents.
| What they do | Configure agents, grant permissions, monitor what they do, handle failures and improve them over time. Somewhere between a system administrator and a team manager. |
| Requirements | Deep understanding of a business process, agent platforms, controls and security thinking, and the ability to read action logs. |
| Pay level | Too new for a reliable benchmark. Pay depends heavily on background: IT, business analysis or operations. |
| Who moves into it | Operations managers, business analysts, RPA and automation specialists, IT staff. |
| Demand in 2027 | Growing fast. The more actions agents take, the more someone needs to own them. |
7. AI Enablement and Adoption Lead
Since February 2025 the EU AI Act has required AI literacy for staff who use AI systems. The adoption lead is the person who makes that happen in practice.
| What they do | Drive tool adoption: role-based training, a champions network, usage and value measurement, and turning experiments into standard processes. |
| Requirements | Change management, training, hands-on knowledge of AI tools, and the ability to talk to both leadership and employees. |
| Pay level | Usually in line with learning and digital transformation roles. No reliable public benchmark for the title yet. |
| Who moves into it | Learning managers, digital project managers, HR and internal communications professionals. |
| Demand in 2027 | High. IBM expects 53% of employees to need reskilling in their current role by 2028. |
8. AI Consultant and Strategist
Second on LinkedIn's 2026 list of fastest-growing jobs in the US. Organizations want someone from outside to tell them where to start.
| What they do | Diagnose where AI will create value, build a roadmap, guide pilots and help choose vendors. |
| Requirements | Business and technology experience, exposure to many organizations, presentation skills and close tracking of the market. |
| Pay level | Varies widely: employees of consulting firms are paid by seniority, independents price by project or day rate. No single public benchmark. |
| Who moves into it | Strategy consultants, product managers, technologists with management experience. |
| Demand in 2027 | High, but competition is growing. Proven case studies make the difference. |
9. AI Evaluator and AI Trainer
A model is only as good as its tests. This role checks answers, writes examples and builds evaluation sets. In professional domains such as medicine, law and finance, companies look for subject experts, not just labelers.
| What they do | Build test sets, rate model answers, write training examples and spot bias and errors. |
| Requirements | Precision, judgment, domain knowledge and sometimes the ability to write simple scripts. |
| Pay level | Depends on expertise: general evaluation work pays like QA and content roles, while experts such as doctors and lawyers are paid hourly, at much higher rates. |
| Who moves into it | QA professionals, content specialists, domain experts and analysts. |
| Demand in 2027 | Stable to rising. LinkedIn lists data annotator and content analyst among growing roles. |
9 roles that changed their substance or their name
Here the structure is the old name, the new one, what it was, what changed and what stays. The recurring conclusion: AI did not erase these jobs. It moved the value from execution to judgment, oversight and accountability.
1. Prompt Engineer: now a skill inside every AI role
| What it was | In 2023 it was a standalone, sought-after job: people whose whole work was writing instructions for models. |
| What changed | Models improved and need fewer "magic words". The title has nearly disappeared from job ads, and the skill moved to AI engineers, product managers and anyone building workflows. Today people talk about context engineering: which information, tools and examples the model receives. |
| What stays | The ability to write, test and refine instructions. It is just no longer enough on its own. |
2. Data Scientist: now AI Engineer or Applied Scientist
| What it was | Training models from scratch, statistical analysis and predictive modeling. |
| What changed | In many companies the work moved to connecting existing language models, evaluating and improving them. Pay data reflects the merge: in our comparison, data scientists sit below ML and AI engineers in every country. |
| What stays | Statistics, experimentation and understanding data. These become an advantage in a world of non-deterministic models. |
3. CIO: now Chief Technology, Data and AI Officer
| What it was | Responsible for infrastructure, systems and security. |
| What changed | In some companies the CIO takes on AI, in others they work alongside a CAIO. IBM found 77% of CEOs say technology and human capital roles are converging. Combined titles such as Chief AI and Data Officer are appearing. |
| What stays | Ownership of infrastructure, security and vendors, the foundation of every AI project. |
4. Privacy and Compliance Officer: now Privacy and AI Governance Officer
| What it was | Responsible for databases, consent and data subject requests. |
| What changed | Privacy laws and the EU AI Act added new questions: what goes into AI tools, who trains on the data, and how to tell a customer they are talking to a bot. In many organizations AI governance was added to the privacy team. |
| What stays | Legal understanding and work with regulators. |
5. QA Tester: now AI Quality and Evaluation Engineer
| What it was | Testing software that returns the same output for the same input. |
| What changed | AI systems answer differently each time. Instead of pass or fail tests, teams build evaluation sets, measure accuracy rates and run red teaming. |
| What stays | The testing mindset, automation and an understanding of what can go wrong. |
6. Copywriter: now Content Editor and Strategist
| What it was | Writing copy from scratch. |
| What changed | The first draft comes from AI, and the value moved to the brief, fact-checking, brand voice and distribution. The EU AI Act also requires disclosure when public text is AI-generated without human editing. |
| What stays | Judgment, storytelling and accountability for what gets published. |
7. Customer Service Agent: now Conversation Manager and Bot Supervisor
| What it was | Answering every request, from simple to complex. |
| What changed | The bot handles simple requests, and the agent handles exceptions, corrects the bot and improves scripts. New roles such as conversation designer appeared. |
| What stays | Empathy, solving complex problems and calming an angry customer. |
8. BI Analyst: now AI Analyst or Analytics Engineer
| What it was | Building reports and queries on request. |
| What changed | Managers ask the data in plain language, and the analyst moves to building the infrastructure that makes the answers correct. Pay moved too: data engineers out-earn analysts in every market we checked. |
| What stays | Business understanding and the ability to ask the right question. |
9. HR Manager: now Manager of People and Agents
| What it was | Recruiting, retention, training and culture. |
| What changed | IBM found 59% of CEOs expect the HR leader's influence to grow, because someone has to plan teams of people and agents. The same study expects 29% of employees to need reskilling into a different role by 2028. |
| What stays | Responsibility for people, culture and fairness, including in AI-assisted hiring. |
What this means for your organization
- Do not appoint a CAIO without authority and budget. An AI leader without control over budget and priorities is a title, not a role. It works only with a clear mandate.
- Start with governance and adoption, not only with hiring engineers. The two roles most often missing in the organizations I work with are an AI governance lead and an adoption lead. They are the ones who turn licenses into managed use.
- Open internal reskilling paths. Many new roles fit people already inside: a compliance officer who becomes the AI governance lead, an operations manager who becomes the agent specialist. That is cheaper and faster than hiring.
- Update pay bands and titles. Employees who lead AI without a matching title or pay are the first to leave, especially with the premiums shown above.
- Plan role-based training. AI literacy is a legal requirement under the EU AI Act for any organization serving EU customers.
I have seen many organizations appoint an AI leader and, a year later, struggle to explain what changed. The difference is one thing: whether that person owns a measurable business outcome, or just a topic. Appointing these roles is part of a broader process that I describe in our guide to AI implementation in organizations.
Frequently asked questions
What are the new AI jobs in 2027?
The main new roles are Chief AI Officer, AI engineer, forward deployed engineer, AI product manager, AI governance and risk lead, AI agent specialist, AI enablement and adoption lead, AI consultant, and AI evaluator or trainer.
How much does an AI engineer earn?
According to Levels.fyi data from October 2026, the median total compensation for a machine learning or AI engineer is about $283,000 in the US, $184,000 in Israel, $137,000 in the UK and $100,000 in Germany, converted at ECB rates. Figures lean toward tech companies and include stock and bonus.
Which country pays AI engineers the most?
Among the four hubs we compared, the US pays the most by a wide margin, followed by Israel, the UK and Germany. The US also has the largest premium for AI engineers over regular software engineers, about 43%.
Is prompt engineering still a job?
Rarely as a standalone title. As models improved, the skill was absorbed into AI engineering, product management and anyone who builds AI workflows. The focus moved to context engineering: which information, tools and examples the model gets.
What does a Chief AI Officer do?
A CAIO owns the organization's AI strategy: where to invest, which systems to approve, how to measure business value and how to govern risk. IBM found 76% of organizations have appointed one, up from 26% a year earlier.
Talk to us about AI in your organization
We help companies operating in Israel map their AI systems, build an AI policy, prepare for the EU AI Act and Israeli privacy law, and train their teams. Leave your details and we will get back to you within one business day.
The Israel AI Organization helps companies and public bodies in Israel adopt AI responsibly. Our daily content is published in Hebrew. About us · Read this guide in Hebrew