Where demand for the radar’s domains is heading (8), emerging micro-skills (15) and declining tasks (7). Every figure on this page links to its source, and every rating is declared editorial.
The global picture
170 million
million new jobs expected globally by 2030
92 million
million jobs expected to be displaced in the same period
39%
of workers’ existing skill sets are expected to be transformed or become outdated by 2030
78 million
million net new jobs expected, as projections create more than they displace
Survey-based expectations of employers worldwide, not a measurement of the Saudi public sector. Press a source badge to see its scope and period.
Where demand for each domain is heading to 2030
Each domain carries an editorial rating that combines named-source signals. Hover over a domain or reach it with the keyboard to see its signals and sources.
Show domains that are:8 of 8 shown
A qualitative editorial direction with no numeric values; the lines show direction only, and the list beside it is the text alternative to the chart.
Domains
AI & Data
Rating: Growing stronglyEditorial rating
Every government entity now carries legal obligations for its data: the Personal Data Protection Law and SDAIA’s national strategy moved data literacy and responsible AI from a niche skill to a compliance requirement. The gap is widening between entities that build the capability in-house and those that rely on buying it.
Signals and their sources
▸AI and big data rank #1 among the fastest-growing skills to 2030.
▸Enrollments from Saudi Arabia in generative-AI courses on a global learning platform grew by about +165% in a year.
▸Globally, the same platform reports more than 45 enrollments per minute in generative-AI content, counted together with a second learning platform’s data.
▸The Personal Data Protection Law in force (2024) and SDAIA’s strategy make data governance a compliance requirement.
Every government entity now carries legal obligations for its data: the Personal Data Protection Law and SDAIA’s national strategy moved data literacy and responsible AI from a niche skill to a compliance requirement. The gap is widening between entities that build the capability in-house and those that rely on buying it.
Signals and their sources
▸AI and big data rank #1 among the fastest-growing skills to 2030.
▸Enrollments from Saudi Arabia in generative-AI courses on a global learning platform grew by about +165% in a year.
▸Globally, the same platform reports more than 45 enrollments per minute in generative-AI content, counted together with a second learning platform’s data.
▸The Personal Data Protection Law in force (2024) and SDAIA’s strategy make data governance a compliance requirement.
Binding controls from the National Cybersecurity Authority make security awareness, identity management and privacy regulatory requirements for every entity, not technical options. Incidents aimed at the public sector keep demand growing strongly to at least 2030.
Signals and their sources
▸Networks and cybersecurity rank #2 among the fastest-growing skills to 2030.
▸Enrollments from Saudi Arabia in cybersecurity courses on a global learning platform grew by about +61% in a year.
▸The National Cybersecurity Authority’s controls (ECC, OTCC) are binding on government entities and critical infrastructure.
▸An international study put the global cyber workforce gap at about 4.8 million people in 2024.
The government cloud wave continues but has moved from decision to operations: demand grows for cloud models, interoperability and digital identity more than for building from scratch. Rated “growing” rather than “growing strongly” because entities mostly buy cloud as a service, so the needed skill is governance and vendor management, not running servers.
Signals and their sources
▸60% of employers see broadening digital access as the most transformative trend for their business.
▸The cloud-first policy and global cloud regions arriving in the Kingdom pull government workloads to the cloud.
▸The national single sign-on platform (Nafath) for digital identity and the Tawakkalna app for government services are now part of everyday operational infrastructure.
Government service quality is now measured and published, turning beneficiary research, journey design and accessibility into criteria required by the Digital Government Authority. Growth is steady and driven by regulation rather than by a technology wave.
Signals and their sources
▸Digital-government standards require beneficiary-centred design and usability evidence.
▸Creative thinking is in the top 5 rising skills in employers’ expectations.
▸Beneficiary-experience measurement programmes embed journey mapping and usability testing in entities’ work.
Every new digital regulation (data protection, AI, open data) creates demand for people who translate legal text into operational practice inside the entity. Digital-legislation awareness rests on judgment under regulatory ambiguity, which makes it hard to automate.
Signals and their sources
▸The Personal Data Protection Law in force (September 2024), open-data requirements and AI-governance frameworks widen the regulatory scope.
▸63% of employers name skill gaps a major barrier to transformation, the biggest barrier in the report.
▸Technology-procurement reform is a stated government priority.
Precision requires a distinction: the agile job title has declined globally, but the practice itself (product lifecycle, user stories, benefits measurement) is now the usual way government digital services are built. Demand is therefore shifting: steady for deep practice, low for formal certificates alone.
Signals and their sources
▸The agile job title has softened globally since 2023 while the practice continues inside teams.
AI agents, digital twins and the Internet of Things moved in 2025–2026 from demos to actual government operations (smart cities, project monitoring, proactive services). This is the radar’s fastest-changing domain, so it needs a content review every six months.
Signals and their sources
▸A global learning platform says learner interest in AI agents reveals growing demand.
▸58% of employers expect robots and autonomous systems to transform their business.
▸National programmes in smart cities, satellite services and digital credentials widen the need for these skills.
The faster automation moves, the more valuable those who lead people through it become: change management, influencing non-technical stakeholders and data storytelling are what turn technology investment into real results. Demand grows steadily, and these skills are less exposed to automation than most.
Signals and their sources
▸About 7 in 10 employers see analytical thinking as a core skill, and resilience, leadership and social influence rank among the top core skills.
▸Enrollments in critical-thinking courses on a global learning platform grew by about +107% in a year.
▸Vision 2030 realization programmes embed change management and data storytelling in entities.
Not degree programmes: specific skills emerging inside existing jobs, each with a sourced demand signal and a practical first step.
15 of 15 shown
Prompt engineering
AI & Data
Structuring instructions to language models for accurate, repeatable outputs in work contexts.
Demand signal: It rides the generative-AI wave: more than 45 enrollments per minute worldwide across two learning platforms combined, and enrollments from Saudi Arabia up +165%.
In government work: Entity staff use approved language tools for drafting and summarizing; a well-built prompt is the difference between dependable output and a costly hallucination.
First step: A short applied course, then build a prompt library for your unit’s recurring tasks.
Designing and supervising chains of AI agents executing multi-step tasks within clear permission boundaries.
Demand signal: AI-agent skills show growing demand on a global learning platform, and 86% of employers expect AI to transform their business.
In government work: Proactive services and cross-department case handling will rely on agents; entities need people who design permission boundaries and human checkpoints.
First step: Run one agent on a low-sensitivity internal task (such as meeting summaries) with a full human-review log.
Auditing generative-model output: spotting hallucination, checking sources, judging when confidence is unwarranted.
Demand signal: Critical-thinking enrollments grew +107% on a global learning platform, a sign of the need for verification skills.
In government work: An undetected error in an official letter or regulatory analysis costs more than in commercial work; verification is the last line of defence before any AI output is adopted.
First step: Adopt the rule “every AI output has a named human reviewer”, and train the team on hallucination checklists.
Managing data quality, definitions, lineage and access rights for data as a documented institutional asset.
Demand signal: AI and big data rank #1 among the fastest-growing skills, and the Personal Data Protection Law requires named data owners inside entities.
In government work: Data exchange between entities stalls without documented definitions and quality; the data steward makes the entity’s data safely shareable.
First step: Start with a lean catalogue of your entity’s most important datasets: definition, owner, quality and sensitivity.
Designing and operating identity and access lifecycles (staff, beneficiaries, systems) across the national single sign-on.
Demand signal: The national single sign-on platform (Nafath) made digital identity shared operational infrastructure, and expanding digital services raise the need to manage it.
In government work: Every digital government service starts with “who is this party?”; sound identity management prevents fraud and shortens journeys.
First step: Review your entity’s current access grants (who holds which right) before any technical certificate.
Systematic pre-build analysis of what could attack a system: assets, attackers, likely paths and mitigations.
Demand signal: Networks and cybersecurity rank #2 among the fastest-growing skills, and the National Cybersecurity Authority’s controls require security by design.
In government work: Government services are standing targets; early threat modeling is far cheaper than remediating a breach of a live service beneficiaries depend on.
First step: Learn one method (STRIDE) and apply it to one of your entity’s digital services in a half-day workshop.
Writing and designing digital content that reaches people with disabilities and older people: plain language, clear contrast, screen-reader support and sign interpretation where needed.
Demand signal: Digital government service standards include accessibility requirements, and mandating them is a documented direction in reference countries.
In government work: A government service must serve everyone; inaccessible content locks whole groups out of services they are entitled to.
First step: Test one service page with a screen reader, fix its most visible issues, then generalize the rule.
Operating and reading live digital replicas of physical assets (buildings, networks, cities) to simulate decisions before execution.
Demand signal: Digital twins are starting to appear in city and infrastructure projects (editorial reading), and 58% of employers expect autonomous systems to transform their business.
In government work: Municipalities and operators need people who can read a twin for maintenance and planning decisions, not only people who build one.
First step: Start with GIS fundamentals, then one twin tool on a single small asset.
Recognizing and verifying machine-generated content (deepfakes, generated voice), and understanding the limits of audio-visual evidence.
Demand signal: The spread of media-generation tools is a documented global trend, and it makes content verification a basic skill for every employee.
In government work: Fabricated rumours about services or officials spread faster than corrections; entities need fast, disciplined detection and response.
First step: Train the communications team on available deepfake-detection tools and on an escalation procedure for confirmed cases.
Enough understanding of renewables, storage, carbon capture and water efficiency to evaluate proposals and track indicators.
Demand signal: Environmental stewardship entered the top 10 fastest-growing skills for the first time, and the Kingdom’s target of about 50% of electricity from renewables by 2030 creates a government need.
In government work: Entities face sustainability reporting and green procurement; climate literacy turns compliance from paperwork into better decisions.
First step: Learn to read a simplified carbon-footprint report for one of your entity’s facilities.
Dividing work between people and models inside procedures: what is automated, what is augmented, what stays human, with review checkpoints designed in.
Demand signal: 77% of employers plan to reskill their workforce to work alongside AI.
In government work: Re-engineering service procedures around AI without designed checkpoints produces either risky automation or token use with no value.
First step: Map one procedure in your unit and classify each step: automate, augment or human work.
Building automated workflows (approvals, notifications, form wiring) with visual tools and no code, with governance that prevents automation sprawl.
Demand signal: AI-agent and automation skills show growing demand on a global learning platform, and no-code tools widen who inside a unit can automate.
In government work: Manual approval queues are among the quickest digital-transformation wins, but ungoverned automation creates shadow systems outside oversight.
First step: Automate one recurring paper procedure (such as an internal leave request) with your entity’s approved tool.
Analyzing spatial data (maps, satellite imagery, administrative boundaries) to support planning and field-service decisions.
Demand signal: Earth observation, digital twins and the national space programme widen the need to read spatial data.
In government work: Municipal, transport and environmental decisions are spatial by nature; analysing “where” improves service distribution and gap detection.
First step: Start with an open mapping tool (QGIS) and plot one service’s beneficiary distribution.
Ethical judgment over behavioural-data use (interactions, patterns, behaviour prediction): consent, fairness and avoiding manipulation.
Demand signal: Responsible AI is a stated priority of the national data and AI strategy, and expanding government recommendation systems require documented ethical judgment.
In government work: A service that “predicts” beneficiary behaviour can improve with behavioural data or slide into discrimination and manipulation; the dividing line is a skill, not a statute alone.
First step: Read the nationally published AI ethics principles and apply their checklist to one project.
The editorial rule: decline hits tasks more than jobs (ILO, 2025); each task shows what remains of it and a suggested shift.
Task
Signal
What declines and what remains
Suggested shift
Manual data entry
Data-entry clerks rank #3 among the fastest-declining jobs worldwide, and clerical work tops the ILO exposure ranking for generative AI (Most exposed group).
The task declines, not every job containing it: those who move to data-quality assurance stay in demand; the difference is between entering data and guaranteeing it.
Short learning in data analysis and quality, then a move into data-quality assurance tasks inside the entity.
Generic (non-specialized) translation
Machine translation and language models now cover routine translation at usable quality, and written-language tasks are among those most exposed to generative AI according to the ILO (2025).
Legal, medical and conference interpreting stay human because of accountability; what declines is generic text, not the translation profession.
Specialize (legal or technical) or move to post-editing machine translation: reviewing and signing off output instead of producing it.
Scripted first-line support
Language-model service assistants are among the clearest generative-AI deployments in live operation, and 86% of employers expect AI to transform their business.
Demand moves to second-line support (complex cases) and beneficiary-experience management: prepared replies are automated; empathy in difficult cases is not.
Beneficiary-experience management and complaint analysis: from answering complaints to preventing them.
Routine bookkeeping
Accounting and bookkeeping clerks rank in the top 10 fastest-declining jobs, and e-invoicing in the Kingdom automates entries at the source.
Accounting as a profession is shifting, not disappearing: analysis and compliance grow while manual entry shrinks.
E-invoicing compliance and financial analysis for the entity, through short credentials available nationally.
Routine administrative and secretarial tasks
Administrative assistants and executive secretaries are among the largest declines in absolute numbers, and scheduling, minutes and template correspondence are among the first tasks language models automate.
Those who run a leadership office (stakeholder coordination, confidentiality, judging priorities) stay; what declines is booking meetings and drafting template letters.
Digital-transformation office management or project coordination: proximity to leadership plus a digital skill.
Template-based graphic production
Graphic designers sit just outside the top 10 fastest-declining jobs for the first time, and image-generation tools produce social and campaign templates in seconds.
Strategic visual identity and user experience are growing; what declines is template execution, not design as thinking.
User-experience design for Arabic-first government digital services.
First drafts and template reports
Language models produce good first drafts of reports and letters, routine writing tasks are among the most exposed according to the ILO (2025), and 40% of employers expect to reduce staff where tasks can be automated.
Value moved from “writing the draft” to “owning accuracy, tone and accountability”: the editor who verifies and signs is worth more than the drafter.
Editing and AI-output verification (a skill in the section above): from production to sign-off.