Leaders & teams
The leader brief
5 shifts in government work between 2026 and 2030, 3 decisions you own this year, and 10 questions for your team. One page you can read in 10 minutes and take to your leadership meeting.
An independent reading — not an official position of any entityLast reviewed:
Employers expect 39% of workers’ existing skill sets to be transformed or become outdated between 2025 and 2030.
5 shifts in government work, 2026–2030
1. Generative AI in everyday casework
What changes: Drafts of letters, reports and replies to beneficiaries start in a generative AI tool, and staff time moves from writing to reviewing and checking. The question becomes: who reviewed the output, and what data did it rely on?
- The skill it asks for
- Responsible use of generative AI: framing the request, checking the output and protecting the data put in. AI & DataCybersecurity
- The decision you need to make this year
- Decide where generative AI may and may not be used in your entity, and who reviews every output before it leaves.
- A signal from a published source:
- The ILO estimates that about 1 in 4 jobs worldwide are in occupations with some degree of exposure to generative AI.
- What does not change
- Accountability for a decision, and the signature on it, stays with the employee and the entity, not the tool.
2. Decisions based on data
What changes: Every department is expected to back its decisions with trusted data: dashboards, published indicators and data shared across entities. A decision without evidence becomes harder to defend.
- The skill it asks for
- Reading and governing data: knowing where a figure comes from, its quality and limits, and having an owner for every dataset. AI & DataLeadership
- The decision you need to make this year
- Name an owner for every dataset your entity’s decisions rely on, and ask that every figure in leadership briefings cites its source.
- A signal from a published source:
- The OECD Digital Government Policy Framework names a “data-driven public sector” as one of 6 dimensions of a digital government.
- What does not change
- Data informs a decision but does not make it; weighing interests and values stays the leader’s job.
3. Services that are measured and published
What changes: Government services are compared in public on satisfaction, ease and time to complete, at home and internationally. The beneficiary’s experience has become a standard the service owner answers for.
- The skill it asks for
- Designing services around the beneficiary: user research, journey mapping, accessibility, and measuring a service’s success after launch. Service DesignAgile Delivery
- The decision you need to make this year
- Pick 3 core services, and give each a named owner and a published indicator you review with them every quarter.
- A signal from a published source:
- The UN E-Government Survey 2024 assesses digital government in 193 countries and publishes the rankings; Saudi Arabia ranked 6th globally on the E-Government Development Index.
- What does not change
- A government service is owed to everyone; people who cannot use the digital channel keep an equivalent alternative.
4. Cybersecurity and privacy as a legal duty
What changes: Protecting data and systems is no longer a technical matter left to one department; it is a legal duty leaders answer for. Every new service needs a privacy assessment and security controls before launch.
- The skill it asks for
- Security awareness, privacy and personal data protection, security by design, and awareness of digital legislation. CybersecurityPolicy & Governance
- The decision you need to make this year
- Ask for a register of services that process personal data, who reviewed each for compliance, and what remains open with a closing date.
- A signal from a published source:
- The Personal Data Protection Law came into force in September 2023, and the compliance grace period for organisations ended in September 2024.
- What does not change
- Trust is built on candour: naming the risks not yet closed beats general reassurance.
5. Product-based delivery
What changes: Entities move from projects that end at launch to digital products with a stable team that keeps improving them. Progress is judged by what actually works for the beneficiary, not by estimated completion percentages.
- The skill it asks for
- Agile ways of working, the product lifecycle, benefits measurement, and technology procurement that pays for outcomes. Agile DeliveryPolicy & Governance
- The decision you need to make this year
- Turn one digital project this year into a product with a stable team, declared stop conditions and a benefit indicator measured after launch.
- A signal from a published source:
- The same framework lists “user-driven” and “government as a platform” among its six dimensions: services built around user needs and improved continuously.
- What does not change
- Financial discipline and accountability for public money stay the same; what changes is when decisions are made, not their standards.
3 decisions you own this year
- 1
Draw the boundaries for generative AI in your entity: where it is allowed, where it is not, and who reviews.
Staff are using these tools now; without a rule, use does not stop but goes out of sight, leaving data and decisions unprotected.
SDAIA’s AI Ethics Principles set out 7 principles, including accountability and responsibility, and transparency and explainability.
Answers shift14AI & Data
- 2
Name an owner for every dataset your decisions rest on, and ask for the source of every figure in leadership briefings.
A decision built on a figure with no owner or source is hard to defend before a board or an oversight body.
The National Data Management Office standards require government entities to govern their data, with clear roles for data ownership and quality.
Answers shift2AI & Data
- 3
Turn one digital project into a product with a stable team and a published service indicator, and measure the result before scaling.
One well-run pilot teaches the entity the new way of working faster than a comprehensive plan on paper.
Answers shift35Agile Delivery
10 questions to ask your CIO and your HR director
For your CIO
Which generative AI tools are our staff actually using today, and what data goes into them?
A reassuring answer: A list of tools and a classification of what data may go in, not “we use nothing”.
Which essential cybersecurity controls have we not yet closed, and when will we?
A reassuring answer: Named gaps with dates and budget, not general reassurance.
If we wanted to leave our largest cloud supplier, how long would it take and what would it cost?
A reassuring answer: A documented estimate, and an exit plan tried on at least one service.
Which of our digital services have a published satisfaction or completion-time indicator, and who owns improving each one?
A reassuring answer: Real indicators reviewed regularly, and a named owner for each service.
Which datasets do our major decisions rest on, and who owns each one?
A reassuring answer: Named owners, and a known quality level for each dataset.
For your HR director
Which jobs in our entity are changing most because of generative AI, and what is our plan for the people in them?
A reassuring answer: Specific job families with a plan for each, not one general programme for everyone.
What digital development have our leaders themselves had this year, and in which domains?
A reassuring answer: Leaders inside the development plan, not outside it.
How do we know training changed how work is done, rather than just adding certificates?
A reassuring answer: Real work examples that changed after training, told by managers rather than trainers.
Do we build scarce digital skills, hire them or borrow them, and on what basis do we decide?
A reassuring answer: A clear criterion for each option, with its cost and timeframe.
How do we make sure skills assessment is used for development only, not for performance appraisal or ranking?
A reassuring answer: A written separation of development from performance, and results that are not collected by name.
Your next step
Sources
The figures and signals above are registered in the evidence log with their source and confidence level.
- World Economic Forum — The Future of Jobs Report 2025 (2025)
- International Labour Organization — Generative AI and Jobs: A Refined Global Index of Occupational Exposure (2025)
- OECD — The OECD Digital Government Policy Framework: Six Dimensions of a Digital Government (2020)
- United Nations DESA — UN E-Government Survey 2024 (2024)
- Saudi Data & AI Authority (SDAIA) — Personal Data Protection Law (PDPL) (2023)
- OECD — The OECD Digital Government Policy Framework: Six Dimensions of a Digital Government (2020)
- Saudi Data & AI Authority (SDAIA) — AI Ethics Principles (2023)
- National Data Management Office (NDMO) — Data Management and Personal Data Protection Standards
An independent reading — not an official position of any entity · Last reviewed: 22 October 2026