Give each role the AI capability it actually needs.

One-size-fits-all AI training creates one-size-fits-none capability.

Generic training often produces content that is irrelevant to some roles, too detailed for others, too shallow for leaders and managers, and weakly connected to the workflows people actually perform each day.

It can also create confusion about responsibilities. A person may leave a session understanding what AI can do but still not know when to use it, what review is needed, which tasks remain human-led, or how the firm expects it to be handled in practice.

Role-based learning starts from the question: “What does this person need to do differently or better?”

Leaders need enough AI understanding to make good business decisions.

Managing partners and senior leaders may need capability around where AI can create value, the service and revenue implications, capacity implications, investment priorities, governance boundaries, adoption barriers, operating-model choices, the difference between specialist capability and internal capability, and how released capacity should be used.

They do not need to become technical AI specialists. Leadership capability is about making informed organisational choices.

Partners need to understand both opportunity and consequence.

Partners and directors need practical understanding of how AI can support professional work, how AI-supported outputs should be reviewed, where professional judgement must remain human, how AI can support client advisory work, where new opportunity may arise, and where accountability for client-facing work still sits.

Good partner learning is commercially and professionally grounded. It helps them assess opportunity without losing sight of quality, trust or professional responsibility.

Managers sit at the centre of practical adoption.

Managers often translate broad strategy into daily delivery. They may need capability around workflow redesign, first-line quality control, reviewing AI-supported work, identifying exceptions, coaching staff, deciding when to escalate, maintaining consistency, implementing new ways of working and identifying where automation or internal AI support can remove friction.

Managers may become one of the most important leverage points in firm-wide AI adoption because they connect strategy to daily practice.

Professional staff need practical skill and professional discipline.

Professional staff often need learning around approved AI tools, useful prompting or instruction, information handling, understanding output limitations, reviewing responses, recognising uncertainty, using AI in defined workflows, escalation and knowing when not to use AI.

Prompting is not the primary capability. The underlying skill is safe, useful use within a professional workflow, with proper review and judgement.

Early-career staff need AI capability without skipping professional development.

Graduates and early-career staff can benefit from AI support for research, organisation, first-pass drafting, summarisation, structured preparation and learning support. But a firm should not allow AI to replace the development of accounting knowledge, critical thinking, professional scepticism, judgement, client understanding or analytical capability.

AI should accelerate learning rather than remove the experiences that create professional judgement.

Some of the strongest AI opportunities may sit outside client delivery.

Operations and support staff often work across administration, internal knowledge, document handling, scheduling, workflow coordination, communication, onboarding, internal reporting and process support. These teams may need strong practical capability around process improvement and AI-enabled workflow design.

The objective is not simply to give them more tools. It is to help them improve how work actually flows through the firm.

Not everyone needs the same technical depth.

Foundational users may need safe practical use, tool basics, review capability and information handling. Power users and champions may need more advanced workflow design, agent configuration, testing and structured instruction. Leaders may need commercial and governance understanding rather than technical configuration.

The important point is not to create rigid user tiers. It is to vary depth appropriately without assuming that more technical detail is always the most important capability.

Different roles carry different AI responsibilities.

Leaders set boundaries and accountability. Managers enforce process and review. Staff follow approved practices and escalate uncertainty. Specialists or champions may test and improve controlled use cases. Partners retain responsibility for consequential professional work.

Governance should make responsibilities understandable at the point of work rather than only in policy documents.

The fastest path from learning to capability is relevance.

Role-based learning should use examples drawn from tax and compliance workflows, client intake, review, advisory preparation, reporting, forecasting, meetings, client communication, internal operations and management information.

Generic examples may teach the tool. Relevant examples teach the work.

People learn governance better when it is connected to real decisions.

Scenarios such as “Can this information be entered?”, “Does this output need review?”, “Is this appropriate for a client?”, “What happens if the AI is uncertain?”, “Should this use case be automated?” and “Who is accountable for the final result?” help make governance practical rather than abstract.

Governance becomes behaviour when it is tied to decision-making in the workflow.

Capability can be built in smaller, relevant pieces.

A role-based learning model may combine short foundational sessions, workshops, practical exercises, real use cases, guided implementation, coaching, curated external courses, internal champions, reusable resources and workflow-specific learning.

This makes the approach far more flexible than assuming that every firm needs a long formal training programme for every role.

Assess before prescribing.

The best learning plan should reflect the current level of capability, the role, the workflow, the stage of adoption, the governance responsibility, the service priorities and the business goals of the firm.

Explore AI Capability Assessment

Some capability gaps are not learning problems.

A weak learning plan is not the only reason adoption may stall. Some gaps require clearer workflow design, stronger systems, better leadership, missing governance, lack of ownership, inconsistent process or the absence of real use cases.

In those situations, the right intervention may be operational or strategic rather than educational.

Use the best learning source for the need.

ScaleEnabler may help identify learning requirements, recommend external programs, curate resources, connect learning to workflows and supplement training with practical implementation support.

A firm does not need to depend on a single provider to build capability. It needs the right combination of learning and practical implementation.

Explore recommended training

Learning should quickly move into real work.

Knowledge becomes useful capability when staff apply it to actual tasks, actual workflows and real client situations. The goal is not years of learning without practice. It is better work carried out with clearer judgement and safer process.

Explore practical AI adoption

People often need help when the real work starts.

Coaching can support applying learning, reviewing outputs, improving instructions, handling exceptions, building confidence and sharing lessons across teams. This is particularly important when AI use becomes embedded in day-to-day work.

Explore coaching & support

Build capability from the work outward.

Role-based learning does not need to start from a blank curriculum.

ScaleEnabler can bring reusable structures such as role capability patterns, accounting-firm use cases, governance scenarios, workflow examples, agent-use patterns, assessment structures, practical exercises, adoption methods and training evaluation questions.

Reusable structure creates efficiency. Firm-specific application creates relevance.

Role-based capability is one part of firm-wide AI maturity.

Strong role-based learning contributes to practical adoption, governance, workflow integration, leadership capability and continuous improvement. But it does not replace those other dimensions.

Explore AI Maturity