Frameworks
AI frameworks for Saudi Arabia: SDAIA, PDPL, NCA and global standards.
Boards, auditors and regulators all ask: measured against what? We map every engagement to the frameworks they already use, from SDAIA and NCA to NIST and ISO.
Comparable. Auditable. Fundable.
We reference and apply these frameworks. Their publishers own them, and we do not speak for them.
How we use them
Every stage has its frameworks. Every gate has its evidence.
Frameworks are not a separate workstream. They shape what we look at in each stage of The Live Path, and what has to be true before each gate opens.
- 01Discover
- 02Design
- 03Lead the build
- 04Run & transfer
Saudi frameworks 8
SDAIA Adoption Framework Live Path stages: Discover, Design, Lead the build, Run & transfer. Details
Saudi Data and AI Authority (SDAIA) · 2024, updated 2025
The national reference for adopting AI, from strategy to measured impact.
How we apply it: In Discover we assess you against the seven components. Each initiative on the roadmap is tagged to the components it moves. The pilot-to-production KPI becomes a gate metric, and Run & transfer reports against Impact.
Applied in: AI strategy & value
SDAIA AI Ethics Principles Live Path stages: Design, Lead the build. Details
SDAIA · 2023 Includes a draft
Seven principles and four risk tiers for anyone building or using AI in the Kingdom.
How we apply it: Every use case gets a risk tier in Design. Principle checks are written into the design sign-off and the Go-Live Gate Checklist. We also prepare you for the draft Responsible AI Policy, labelled as a draft until it is final.
Note: The Responsible AI Policy is still a draft; its public consultation ran from April to May 2026. Check its status before relying on it.
Applied in: Governance & regulatory readiness
SDAIA Generative AI Guidelines Live Path stages: Design, Lead the build. Details
SDAIA · 2024
Two editions, for government staff and for the public, on using generative AI tools safely.
How we apply it: We turn the guidelines into a usage policy and an approval process for tools. Then we build the same rules into the assistants we deploy: data filters, review steps and labels.
Applied in: Governance & regulatory readiness
SDAIA AI Risk Framework Live Path stages: Design, Run & transfer. Details
SDAIA · July 2026
A national, repeating risk cycle for AI, built on the ethics principles.
How we apply it: We run the Saudi cycle and the global standards (NIST AI RMF, ISO/IEC 23894) in one risk register. Your team keeps a single record, not three.
Applied in: Governance & regulatory readiness
PDPL Live Path stages: Design, Lead the build. Details
SDAIA (regulator) · In force since September 2023 Includes a draft
The Kingdom's personal data law, with its Implementing Regulation and transfer rules.
How we apply it: We map each personal data flow an AI system touches before design sign-off. If a model or API sits outside the Kingdom, the transfer assessment comes first, and it often decides the hosting.
Note: Draft amendments to the Implementing Regulation closed for consultation on 5 November 2026. Check their status before relying on them.
Applied in: Governance & regulatory readiness
Sources:- Umm Al-Qura gazette, amendment to the Personal Data Protection Law (2023)
- Umm Al-Qura gazette, Regulation on Personal Data Transfer outside the Kingdom (2024)
- DLA Piper, Data Protection Laws of the World: Saudi Arabia
- Clyde & Co, SDAIA publishes draft amendments to the PDPL regulations (8 October 2026)
NDMO Standards Live Path stages: Discover, Design. Details
National Data Management Office (NDMO), SDAIA · v1.5, 2021
Fifteen data management domains, plus national policies on classification and sharing.
How we apply it: In Discover we assess the data an AI system needs against the relevant domains, and classify it before Design. The classification then drives the hosting choice: on-premises, sovereign cloud or a licensed cloud.
Applied in: Data, platforms & sourcing
Sources:- NDMO, National Data Management and Personal Data Protection Standards v1.5 (copy on Qanoniah)
- Ministry of Education, Data Governance Policies v1.1 (NDMO policies on classification, sharing, FOI and open data)
- NCA, Essential Cybersecurity Controls ECC-2:2024 (English), including Appendix C on data localisation moving to NDMO
NCA ECC and CCC Live Path stages: Design, Lead the build, Run & transfer. Details
National Cybersecurity Authority (NCA) · ECC-2:2024 and CCC-2:2024
The national minimum cybersecurity controls, and their extension for cloud.
How we apply it: We map the AI system and its suppliers to the relevant controls in Design, then verify them at the go-live gate. Model providers and APIs are reviewed as third parties.
Applied in: Governance & regulatory readiness
DGA standards Live Path stages: Discover, Design. Details
Digital Government Authority (DGA) · Basic Standards v3.0 (2024); readiness guide 2026
How government entities are measured on digital transformation and emerging technology.
How we apply it: For government and semi-government clients, we link each AI initiative to the Qiyas standards and readiness criteria it improves, and document the architecture in NORA terms.
Applied in: AI strategy & value
Sources:- DGA, Digital Transformation Basic Standards v3.0 (2024, Arabic)
- DGA, Qiyas digital transformation measurement, 2025 results
- DGA, Emerging Technology Adoption Readiness guide 2026
- DGA, Guideline for National Overall Reference Architecture (NORA) v1.0 (2024)
- DGA, Enterprise Architecture Establishment for Government Entities (2022)
Strategy and maturity 5
McKinsey Rewired Live Path stages: Discover, Run & transfer. Details
McKinsey & Company · 2023; 2nd edition 2026
Six enterprise capabilities that McKinsey links to successful digital and AI transformations.
How we apply it: We use it as a completeness check in Discover: can each initiative land with today's talent, operating model and data? Adoption gets its own budget line in the business case.
Note: Rewired is McKinsey & Company's framework. We reference it. We are not affiliated with McKinsey.
Applied in: AI strategy & value
BCG 10-20-70 Live Path stages: Discover. Details
Boston Consulting Group (BCG) · 2024
Ten percent algorithms, twenty percent technology and data, seventy percent people and processes.
How we apply it: We use it as a budget test. If the people and process share is missing, the business case is not ready. Our operations and adoption practices carry the seventy.
Applied in: AI strategy & value
Three Horizons Live Path stages: Discover. Details
McKinsey & Company (Baghai, Coley, White) · 1999
Fund today's core, tomorrow's businesses and future options at the same time.
How we apply it: In Discover we place the ranked roadmap on the three horizons, so the board sees the balance between efficiency in core operations and new AI-enabled services.
Applied in: AI strategy & value
MIT CISR AI Maturity Live Path stages: Discover, Run & transfer. Details
MIT Sloan Center for Information Systems Research (CISR) · 2024
Four stages of enterprise AI maturity, linked to financial performance.
How we apply it: We use it as a plain-language reading for boards, next to SDAIA's maturity levels: where you are, and which stage the roadmap should reach.
Applied in: AI strategy & value
Gartner AI Maturity Live Path stages: Discover. Details
Gartner · Toolkit, 2024
A maturity assessment across seven pillars, from strategy to people and culture.
How we apply it: Many clients already hold a Gartner assessment. We read it alongside the SDAIA baseline, so the two views agree instead of competing.
Note: Gartner's full model is paid research. We name it and work with assessments our clients already hold.
Applied in: AI strategy & value
Governance and standards 5
NIST AI RMF Live Path stages: Design, Lead the build, Run & transfer. Details
US National Institute of Standards and Technology (NIST) · 1.0 (2023); Generative AI Profile (2024)
Four functions for managing AI risk, with a profile for generative AI.
How we apply it: Govern and Map shape Design. Measure becomes the test plan in Lead the build, using the generative AI risks. Manage carries into Run & transfer.
Applied in: Governance & regulatory readiness
ISO/IEC 42001 Live Path stages: Design, Run & transfer. Details
ISO and IEC · 2023
The international standard for running AI under a management system.
How we apply it: We design the AI office and its policies so they can grow into a 42001 management system if you decide to certify later. Certification itself comes from an accredited body.
Note: We do not certify. Accredited certification bodies do.
Applied in: Governance & regulatory readiness
ISO/IEC 23894 Live Path stages: Design. Details
ISO and IEC · 2023
Guidance for fitting AI risk into the risk process you already run.
How we apply it: Many entities already run enterprise risk management in the ISO 31000 style. We set AI risk criteria that slot into that register.
Applied in: Governance & regulatory readiness
OECD AI Principles Live Path stages: Design. Details
OECD · 2019, updated 2024
Five principles for trustworthy AI, adopted by governments around the world.
How we apply it: A reference point when your AI policy must read well for international partners, investors or subsidiaries.
Applied in: Governance & regulatory readiness
EU AI Act Live Path stages: Design. Details
European Union · Regulation (EU) 2024/1689, amended 2026
The EU's risk-based AI law. Relevant when your AI output is used in the EU.
How we apply it: Only for GCC clients whose AI output reaches the EU, through customers, subsidiaries or partners. We check the tier for each use case and note which dates apply.
Applied in: Governance & regulatory readiness
Delivery and data 4
CRISP-DM Live Path stages: Design, Lead the build. Details
CRISP-DM consortium; MLOps levels by Google Cloud · CRISP-DM 1.0 (2000)
Six phases for data and model work, and three levels of MLOps automation.
How we apply it: It is the inner loop for model work in Design and Lead the build. The target MLOps level is a go-live criterion, set at what your team can run.
Applied in: Operations re-engineering
DAMA-DMBOK Live Path stages: Discover, Design. Details
DAMA International · 2nd edition, revised 2024
Eleven data management knowledge areas, with governance at the centre.
How we apply it: We use it as a common language with your data office, and cross-reference our findings to the NDMO domains.
Note: DAMA-DMBOK is a registered mark of DAMA International.
Applied in: Data, platforms & sourcing
TOGAF Live Path stages: Design. Details
The Open Group · 10th Edition
A method for enterprise architecture, from vision to change management.
How we apply it: When your enterprise architecture team works in TOGAF or NORA, we deliver AI solution architecture in those terms.
Note: TOGAF is a registered trademark of The Open Group.
Applied in: Data, platforms & sourcing
ITIL Live Path stages: Run & transfer. Details
PeopleCert · ITIL 4; ITIL Version 5 (2026)
The common practice for running IT services, now updated for AI and cloud.
How we apply it: In Run & transfer we hand the system to your service desk and operations teams, inside the ITIL processes they already run.
Note: ITIL is a registered trademark of PeopleCert.
Applied in: Digital core & ecosystems
What has to be true
- 01 Discover. A ranked roadmap with business cases.
- 02 Design. Design sign-off and a value baseline.
- 03 Lead the build. The Go-Live Gate Checklist, signed.
- 04 Run & transfer. Value signed off by finance, and the next operation chosen.
This mapping is our working method. The publishers do not prescribe it.
The backbone
SDAIA's AI Adoption Framework.
Three pillars and seven components, from strategy to impact. Each of our six practices maps to one or more components, so every recommendation lands on something you can report against.
What is SDAIA's AI Adoption Framework?
SDAIA's AI Adoption Framework is Saudi Arabia's national reference for adopting AI, from strategy to measured impact. It has three pillars and seven components. The four maturity levels belong to the 2024 edition; the 2025 edition has none, and SDAIA is developing a maturity index for government entities.
- Publisher
- The Saudi Data and AI Authority (SDAIA)
- Launched
- September 2024, updated in 2025
- Applies to
- All government and private entities
- Octopus
- References the framework; not affiliated with SDAIA
Directions
Strategy · Governance
We link your roadmap to national priorities and make sure a named person is accountable for AI.
Enablers
Data · Infrastructure · Human capacity
We score your data, platforms and skills in Discover. The roadmap then closes the gaps.
Outcomes
Applications · Impact
We tag each initiative to the components it moves. In Run & transfer, we report against Impact.
Maturity levels (2024 edition)
- Emerging
- Developing
- Proficient
- Advanced
The KPI we track at every gate: Projects moving from pilot to production. The framework names this measure itself.
Prefer to talk it through? Bring your roadmap or RFP, and we will map it to the frameworks with you.
30 minutes with our team. We tell you plainly where we would start.