The directions employers are taking roles
Each direction is a way employers describe a role changing as they adopt AI. The share is the part of all AI-related postings that point that way; a posting can point in more than one.
- AI transformation64%
In 43 of 44 roles. Strongest in Finance Manager, Executive Assistant, Project Manager.
- Data & machine learning24%
In 23 of 44 roles. Strongest in Data Engineer, Data Analyst, Data Architect.
- AI engineering24%
In 16 of 44 roles. Strongest in AI Engineer, DevOps Engineer, Software Engineer.
- AI sales23%
In 9 of 44 roles. Strongest in Sales Development Representative, Business Development Representative, Sales Manager.
- AI customer experience13%
In 14 of 44 roles. Strongest in Customer Success Manager, Account Manager, Account Executive.
- AI product12%
In 13 of 44 roles. Strongest in Product Manager, Product Owner, Product Designer.
- AI marketing7%
In 13 of 44 roles. Strongest in Social Media Manager, Marketing Specialist, Digital Marketing Specialist.
- AI governance6%
In 16 of 44 roles. Strongest in Enterprise Architect, Program Manager, Executive Assistant.
The biggest change in each field
The direction that comes up most within each group of roles.
Engineering & architecture
AI engineering
82% of the field's postings
Data & AI
Data & machine learning
77% of the field's postings
Product & design
AI transformation
75% of the field's postings
Sales & customer success
AI sales
88% of the field's postings
Marketing
AI marketing
100% of the field's postings
Management & operations
AI transformation
100% of the field's postings
Finance, legal & risk
AI transformation
74% of the field's postings
The skills these changes add
How many roles have each skill among the skills their directions add.
- AI Change Management33 of 44 roles
- Change Management32 of 44 roles
- Prompt Engineering30 of 44 roles
- AI Tool Evaluation27 of 44 roles
- AI Ethics21 of 44 roles
- AI Governance20 of 44 roles
- AI Compliance19 of 44 roles
- AI Strategy19 of 44 roles
- AI Risk Management16 of 44 roles
- MLOps16 of 44 roles
The responsibilities that recur
Most responsibilities these changes add are specific to one role; these are the ones that recur across the most roles.
- Conduct AI risk assessments and ensure compliance with emerging AI regulations (8 roles)
- Conduct adversarial testing and red-teaming of AI systems (5 roles)
- Implement monitoring and observability for AI systems to ensure reliability and performance (4 roles)
- Implement security controls for AI models and data pipelines (4 roles)
- Analyze campaign performance using AI-driven analytics to inform strategy (3 roles)
- Conduct AI risk assessments and audits (3 roles)
FAQ
How is AI changing jobs?
Mostly by folding AI into the job rather than replacing it. The most common change is AI transformation, which comes up in 64% of AI-related postings across 43 of 44 roles. Next come data & machine learning (24%), AI engineering (24%) and AI sales (23%).
What new responsibilities is AI adding to jobs?
The ones that recur across the most roles are: “Conduct AI risk assessments and ensure compliance with emerging AI regulations” (8 roles); “Conduct adversarial testing and red-teaming of AI systems” (5 roles). Most added responsibilities are specific to one role, and each role guide lists its own.
Which skills are employers adding because of AI?
Across the directions employers describe, the skills added in the most roles are AI Change Management (33 of 44 roles), Change Management (32 of 44 roles) and Prompt Engineering (30 of 44 roles).
See the detail for your role
Every role guide shows its essential skills, the AI skills to learn next, the common gaps and where the role is heading.
Browse all 44 role guidesBuilt from 57,486 public job descriptions across 44 roles, published from March to September 2026 and deduplicated so that a job listed on several boards or by several agencies counts once. Directions, skills and responsibilities are extracted from each job description. A posting can point in several directions, so the shares add up to more than 100%; each is weighted by the role's number of postings. How we collect and deduplicate postings.