This checklist is based on the AI maturity framework in ThoughtLab’s “The AI-powered investment firm” report and playbook, of which Grant Thornton is a sponsor and contributor. The checklist translates its core principles into seven pragmatic steps for asset and wealth management leaders. It shows leaders how to evolve along the maturity continuum, achieving responsible, measurable progress — boosting growth, efficiency and control without overwhelming their teams.
Introduction
For many asset and wealth management teams, “AI” still feels like a one-size-fits-all buzzword. In reality, each firm needs a level and pace of AI adoption that matches its own goals, capabilities and controls. This checklist — based on the AI maturity framework developed by ThoughtLab, an analytics-driven thought leadership firm — helps you pinpoint where you are on that journey and choose the next practical steps to deliver real value.
The AI maturity framework considers the foundational pillars required for successful AI transformation:
- AI strategy and implementation plan
- Innovation culture
- Use of a modern IT platform
- Right mix of AI talent and skills
- Adoption of advanced AI tools
- Effective data management
- AI governance framework
Regional surveys show that AI adoption is accelerating. For instance, a recent DFSA study found that 52% of financial firms in Dubai’s DIFC are now using AI (up from 33% in 2024), with GenAI usage surging.
However, adoption alone doesn’t guarantee value. Independent industry and analyst studies indicate that ROI remains uneven as many organisations move from pilots to scaling — with returns concentrated among early, well-governed deployments.
In short, many firms are experimenting with AI, but do they have the correct foundations and roadmaps in place to truly capitalise on these tools? And how does your firm compare?
AI maturity framework
Step 1. AI strategy and implementation plan
Create an effective AI strategy and implementation roadmap.
What it looks like in practice:
Your firm develops a business-led AI strategy before implementing any technology roadmap — clarifying ambitions, guardrails and priority use cases. Executives agree on the problems AI will solve, the value expected (revenue, cost, risk) and the risk appetite that guides every decision.
Why it matters:
A shared strategy prevents ‘random acts of AI’ and instils confidence in first moves.
Your next step to an AI strategy:
- Define clear, measurable objectives and success metrics tied to business goals before making any technology decisions.
- Build alignment across the organisation to support adoption of the strategy, objectives and metrics, and ensure key leaders are involved in the decision-making process.
Step 2. Innovation culture
Cultivate an innovation culture that encourages responsible AI experimentation, aligned with the UAE’s forward-leaning approach to AI.
What it looks like in practice:
Leaders normalise controlled experimentation — lessons learned are shared and successes scaled quickly. Pilot teams present results to leadership every quarter; high-value experiments are expanded swiftly, while unproductive ones are retired promptly. In the UAE’s innovation-oriented landscape, this also means working with regulator-approved sandboxes and innovation programmes.
For example, DIFC and ADGM regulators actively support fintech experimentation through initiatives such as the Innovation Testing Licence sandbox, enabling firms to test AI solutions in a controlled environment.
This culture rewards innovation but remains mindful of the firm’s risk limits and the UAE’s regulatory expectations regarding emerging technology.
Why it matters:
Visible and proactive leadership sponsorship accelerates adoption and prevents innovation fatigue.
Your next step to an AI strategy:
- Start small with low-risk AI use cases.
- Automate prospect meeting notes for financial advisors: Capture and summarise key goals and needs, then surface next steps for follow-up.
- Launch a private internal AI assistant: Enable teams to quickly find policies, product documents and answers to routine enquiries — without exposing client data.
- Then, progress to:
- Personalised engagement and investing: Offer real-time portfolio views and recommendations tailored to a client’s goals, risk and life stage.
- Agentic AI for operations: Streamline data aggregation, reconciliation, compliance and reporting with software agents for narrow, pre-approved tasks — such as data aggregation, reconciliation drafts and report prep — with human review for any final action.
- Embedding AI across the business: Expand adoption through improved data foundations, change management and governance.
Step 3. Use of a modern IT platform
Establish a modern, cloud-based IT platform to facilitate AI adoption across systems.
What it looks like in practice:
Core systems are cloud-ready, secure and flexible enough for AI workloads. When workload spikes — for example, during month-end performance calculations or large-scale risk simulations — the cloud automatically adds computing power, then scales back once processing is complete, all within a zero-trust security framework.
Why it matters:
Modern platforms cut downtime, cost and cyber risk — the bottlenecks that stall AI adoption.
Your next step to an AI strategy: Case study
- Scenario: An asset manager lacked cloud-based core systems but wanted to pilot and scale AI.
- Approach: We implemented a cloud automation solution using a large language model to process legal documentation, guided by a capability-mapping framework that linked core LLM capabilities to business functions.
- Result: A 340% ROI driven by faster turnaround, reduced manual effort and improved data accuracy.
Step 4. Right mix of AI talent and skills
Develop or acquire AI talent and skills across the enterprise.
What it looks like in practice:
Your people receive role-specific training so they can use AI tools confidently and responsibly in everyday decisions. Investment, operations and client service teams consult AI dashboards or assistants to test scenarios and validate decisions. In the UAE’s multicultural workforce, training programmes are tailored for a bilingual environment – enabling both English and Arabic speakers to build AI literacy.
Why it matters:
When AI is woven into normal workflows, decisions are made faster, adoption is sustained and specialist data scientists can focus on higher-value innovation.
Your next step to an AI strategy:
Many asset and wealth management teams lack practical fluency in technology use and in making data-literate decisions. An optimal method to upskill your organisation is through role-based, scalable training tied to business objectives and reinforced by clear responsibilities and career paths.
Any upskilling initiative should also account for the UAE’s diverse multilingual talent pool. In fact, new industry programmes are emerging to close the skills gap; for instance, an ‘AI Academy’ was launched to deliver Arabic-language AI education and certifications for professionals in finance and other sectors.
Step 5. Adoption of advanced AI tools
Draw on the latest AI technologies, such as generative and agentic AI, where they add value.
What it looks like in practice:
The firm selects proven, purpose-built solutions that solve real pain points. For example, a ready-made AI text analysis tool can scan prospectuses and research reports in minutes, freeing analysts to focus on developing insights rather than extensive reading. AI models could also be used to screen large datasets for Shariah compliance or flag anomalies in trading patterns, augmenting the team’s capabilities.
Why it matters:
Quick wins build confidence and free up time and budget for larger, longer-term projects that deliver stronger results.
Your next step to an AI strategy:
How to decide — build or buy?
- Start with the pain point: choose the option that solves a defined business problem.
- Check integration and controls: prioritise data security, audit trails and user management in your environment. Ensure any third-party AI vendors or cloud platforms comply with UAE outsourcing guidelines and data protection laws (e.g. UAE PDPL for personal data).
- Compare time to value and total cost: pilot in weeks and track one primary metric (cycle time, error rate or dollars saved).
Shane O’Neill, Consulting Partner, Grant Thornton Ireland, explained how choosing the right tool resulted in clear ROI for one asset manager.
Similar automation opportunities are emerging for UAE asset managers as AI adoption accelerates and regulatory expectations evolve. Our team’s cross-border expertise and local experience in delivering measurable operational improvements can support UAE firms as they expand their use of AI.
Step 6. Effective data management
Build and maintain a system for cleansing, integrating and optimising firmwide data.
What it looks like in practice:
Data is clean, connected and governed before models go live. All critical data resides in one (or seamlessly federated) trusted, well-organised source that automatically feeds reports and AI tools, with built-in alerts that flag quality issues before decisions are made.
Data storage and transfers comply with applicable data sovereignty rules – for example, client personal data is handled in line with UAE federal privacy law and any relevant free zone regulations on cross-border data.
Why it matters:
Trusted data underpins reliable AI outputs and keeps processes audit-ready. Fragmented, inconsistent source data is the most common blocker to AI adoption because it erodes trust in outputs.
Your next step to an AI strategy:
Many firms in the region grapple with siloed or inconsistent data scattered across systems and jurisdictions. A practical first step is to conduct a quick data maturity assessment, then address the highest-impact data quality and integration gaps. Ensure each AI use case is fed with clean, readily available data that fits how teams work.
This approach should also factor in data residency and outsourcing rules – for example, if using a cloud provider, the firm must remain accountable for how data is stored and processed. UAE regulators such as CBUAE, SCA, DFSA and FSRA have jointly emphasised robust data management in their guidelines for adopting new technologies.
“By investing in data cleansing and integration up front, asset managers can avoid the ‘garbage in, garbage out’ pitfall and meet regulators’ expectations for sound data management,” said Marwan Galal, Technology Consulting Director, Grant Thornton UAE.
Step 7. AI governance framework
Install governance policies and structures for the responsible and effective use of AI.
What it looks like in practice:
Policies and oversight keep AI transparent, compliant and in line with the firm’s risk limits. This aligns with national initiatives like the UAE’s National Strategy for AI 2031, which aims to make the country the world’s most AI-ready nation, and the UAE’s AI Ethics Guidelines emphasising principles of fairness, transparency and accountability.
A cross-functional AI governance council reviews every major AI model for clarity and regulatory fit before launch. They ensure each algorithm meets relevant guidelines (for example, alignment with the DFSA/FSRA expectations on algorithmic transparency and the SCA’s outsourcing requirements for tech solutions). The council then tracks real-world performance, data drift and controls throughout the lifecycle.
Why it matters:
Strong governance is the gate that transforms small pilots into enterprise-scale solutions that the board, regulators (SCA, DFSA, FSRA) and clients can trust. In the UAE’s financial sector, regulators have signalled that AI adoption must go hand-in-hand with effective oversight and ethical use.
Your next step to an AI strategy:
- Start small and make governance repeatable.
- Establish an AI ethics committee or working group to review AI projects, define roles and set standards.
- Put a control framework in place with model documentation, audit logging, performance monitoring.
- Ensure there is human review to management exceptions, with a right-sized AI literacy programme as adoption scales firmwide.
Consider these AI questions
Overall, what ROI, if any, are you seeing from your uses of AI? Please consider the full range of benefits and costs of the investment.
Tip: AI-mature firms report large positive ROI and are already piloting or scaling GenAI.
How extensively do you use the following AI technologies across your business?
Tip: Wide use of some or all of these technologies indicates an AI-mature organisation.
Next steps on the AI maturity continuum
Your position on the maturity spectrum isn’t a label — it’s a launch point. Strengthen the pillar that will unlock the greatest business value next, measure ROI early and repeat. Firms that follow this discipline move from isolated pilots to enterprise-wide AI that grows revenue, cuts costs and remains compliant.
In the UAE, this disciplined approach is also key to meeting stakeholder and regulator expectations. Start small, think big and proceed step by step. As you advance, stay aware of industry benchmarks. UAE asset managers are increasingly looking to each other (and global peers) to gauge progress.
Find out how your firm stacks up against peers in ThoughtLab’s “The AI-powered investment firm” report and playbook, in partnership with Grant Thornton.
