Key numbers
Job postings checked
528,697
Mention AI or machine learning
21,640
4.09% of job postings checked
Role groups compared
12
across the archive
Highest role group
34.01%
Software & data
AI mentions by role group
Percent of job postings checked. All bars use a 0–40% scale.
Mentions by role group
Role groups with the highest percentage appear first.
| Role group | Mentions | Checked | Percent |
|---|---|---|---|
| Software & data | 8,925 | 26,244 | 34.01% |
| Product & design | 1,093 | 5,443 | 20.08% |
| IT & security | 2,803 | 24,713 | 11.34% |
| Marketing & sales | 2,776 | 37,173 | 7.47% |
| Finance & risk | 1,174 | 28,178 | 4.17% |
| Engineering & science | 1,644 | 45,023 | 3.65% |
| Other | 75 | 2,770 | 2.71% |
| Business operations | 1,851 | 79,764 | 2.32% |
| Education & public service | 215 | 20,563 | 1.05% |
| Retail & hospitality | 597 | 116,882 | 0.51% |
| Healthcare & life sciences | 400 | 106,873 | 0.37% |
| Construction & field services | 87 | 35,071 | 0.25% |
AI terms used in job postings
Number of job postings mentioning each kind of AI language. One job posting can appear in several rows.
Roles with most AI mentions
Only roles with at least 500 job postings checked are shown.
| Role | Mentions | Checked | Percent |
|---|---|---|---|
| AI & machine learning | 2,107 | 2,775 | 75.93% |
| Mobile engineering | 282 | 679 | 41.53% |
| Full-stack software | 1,650 | 4,395 | 37.54% |
| Data & analytics | 2,712 | 7,366 | 36.82% |
| Cloud & DevOps | 592 | 1,660 | 35.66% |
| Frontend engineering | 252 | 807 | 31.23% |
| Tech leadership | 768 | 2,559 | 30.01% |
| Backend engineering | 720 | 2,447 | 29.42% |
| Product management | 664 | 2,553 | 26.01% |
| Cybersecurity & identity | 804 | 3,673 | 21.89% |
| Sales operations | 203 | 998 | 20.34% |
| Design & creative | 279 | 1,590 | 17.55% |
| Sales engineering | 119 | 699 | 17.02% |
| Hardware & robotics | 559 | 3,904 | 14.32% |
| Media & editorial | 127 | 1,135 | 11.19% |
What job postings ask people to do
These examples show the tasks found while reading job postings, not how often each task appears. Each excerpt comes from a different job posting. Employer names and identifying details have been removed; excerpts are lightly trimmed and punctuation is cleaned up.
| Role group | What job postings say |
|---|---|
| Software & data |
Build AI search and agent systems, connect models to company tools and data, use coding assistants, and test whether outputs are reliable and affordable. “Develop and maintain the AI agents, prompts, retrieval pipelines, and decision logic that power those workflows.” “AI accelerates implementation, but engineers remain responsible for system design, debugging, code review.” “Architect and build MCP servers that expose the enterprise API catalog as a queryable, real-time context, enabling AI agents and developer tools.” “Develop field-relevant benchmarks for agentic tasks in biological design.” “Utilize AI agents to accelerate delivery of complex and large-scale business requirements.” |
| IT & cybersecurity |
Set up approved AI services, manage access and data security, and use AI to help with support, networks, threat detection, and incident response. “Identify repetitive, high-volume SOC workflows and systematically eliminate them through AI-powered triage pipelines, LLM-assisted investigation, and end-to-end automation.” “Assess AI/LLM risk: prompt injection, tool data-exfiltration, over-broad tool access.” “Applying AI/agentic tooling to automate security operations tasks (e.g., ticket triage, anomaly detection, workflow orchestration).” “Design and architect an enterprise website search functionality using advanced Artificial Intelligence and Machine Learning techniques.” “Leveraging automation, scripting, machine learning, and AI to support cybersecurity operations.” |
| Product & design |
Decide where AI belongs in a product, quickly prototype ideas, and design clear approvals, explanations, and handoffs to people. “Use AI coding tools to creatively prototype and test ideas with real users in hours, not weeks.” “Design experiences for our AI and agentic capabilities, turning complex system behavior into flows that feel intuitive and trustworthy.” “Lead the definition and delivery of AI-enabled and agentic workflows, identifying repeatable processes that can be automated or delegated to agents.” “Convert AI/GenAI use cases and business needs into clear user stories, requirements, and acceptance criteria.” “Defining the conversation flows, prompt structures, and behavioral standards that make agentic AI experiences feel trustworthy, capable, and human-centered.” |
| Marketing & sales |
Find and prioritize prospects, personalize outreach, optimize marketing spending, and improve visibility in AI-powered search while reviewing accuracy and tone. “Leverage generative AI tools to draft and iterate on personalized outreach messaging, then apply sound judgment and domain knowledge to refine that content.” “Generative AI for content creation, predictive analytics for targeting and scoring, and AI-powered personalization at scale.” “Getting our content cited inside AI answer engines like ChatGPT, Perplexity, Gemini, and Google AI Overviews.” “Optimize content for visibility and citation across platforms such as Google AI Overviews, ChatGPT, Gemini, Claude, and Perplexity.” “Comfortable leveraging AI tools (e.g., ChatGPT, Gemini, Claude) to improve research, prospecting, proposal creation, and daily sales workflows.” |
| Finance, legal & risk |
Use AI for research, analysis, drafting, reporting, and review—then validate the result and maintain human oversight and audit controls. “Use tools such as Claude or ChatGPT to build automated workflows that streamline financial reporting, variance analysis, and data reconciliation.” “Support technology-assisted workflows on live matters, such as contract analysis, due diligence, document automation, transaction management, and litigation.” “Co-develop AI agents with AI engineers to execute detailed SOX testing by translating SOX testing methodology into specific guidance, decision rules, and examples.” “Perform independent credible challenge of enterprise AI use cases (including agentic AI and LLMs), assessing compliance, control effectiveness, output risk, and customer impact.” “Uses advanced analytics, machine learning, and statistical techniques to design, develop, and optimize Anti-Money Laundering (AML) transaction monitoring solutions.” |
| Engineering & science |
Apply machine learning to simulation, physical design, computer vision, and biological research, while checking that results make scientific sense. “Design, implement, and optimize the algorithms that power our AI computer vision systems.” “Applying machine learning, computer vision, and real-time data processing to automate and accelerate every step from click to delivery.” “Developing Python-led robotics applications, ROS2-based integration, computer vision pipelines, embodied AI workflows and simulation-to-hardware validation.” “Leverage cutting edge AI and ML approaches, including LLM powered agentic workflows and network based methods, to accelerate target triage and automate biological evidence synthesis.” “Applying Artificial Intelligence and Machine Learning to next-generation automation and robotics solutions, including predictive control, fault detection, anomaly detection, predictive maintenance.” |
| Business operations |
Automate recurring office and procurement work, support executives, introduce approved tools, train coworkers, and measure whether the new process helps. “Automate and simplify recurring operational processes—reporting cycles, briefing preparation, meeting rhythms—using GenAI tools to free capacity for higher-value work.” “Leverage generative AI tools as a core partner in the workflow, to automate data synthesis, and optimize end-to-end supply chain strategies.” “Drive a step-change reduction in manual work across all aspects of People Operations through automation, AI agents, self-service, and intelligent workflows.” “Drive AI-enabled process improvements, using tools like ChatGPT and Claude to automate reporting, streamline documentation, and increase alliances team efficiency.” “Adopting modern productivity tools, workflow automation, and Generative AI assistants (e.g., Gemini, Claude, NotebookLM) to streamline daily operations.” |
Method and limitations
Counts show where specific AI and machine-learning terms appear; they do not prove that a job requires AI skills. Role groups describe the kind of job, not the employer’s industry. One job posting can appear in several term rows, and these results do not represent every open job worldwide.