From AI Curiosity to Firmwide Capability

What the AGN AI & Automation Masterclass Reveals About Real AI Adoption in Mid-Sized Firms.

AI adoption isn’t about choosing the right tools—it’s about building the right capabilities. Discover the key insights, leadership lessons, and practical recommendations from the AGN AI & Automation Masterclass to help your firm move from experimentation to firmwide impact.

Action Now: Practical Steps for AI Adoption

Treat AI as an operating model issue, not a technology initiative. Assign senior ownership to AI-enabled processes in the same way as core service lines or regulatory obligations.

Prioritise a small number of material use cases. Focus on processes that directly affect client experience, speed-to-revenue, risk, or staff capacity rather than spreading effort across multiple pilots.

Get data readiness right early. Invest upfront in data structure, integration, and quality; weak data is the most common hidden blocker to AI progress.

Build change leadership into every AI initiative. Training, communication, and behavioural change must be planned from the outset, not added once resistance appears.

Define human-in-the-loop controls explicitly. Be clear about where professional judgement sits, how exceptions are handled, and who is accountable for outcomes.

Use NextGen capability deliberately. Empower emerging leaders to design and deliver AI initiatives, but anchor success through visible partner sponsorship and enforcement.

Measure adoption, not experimentation. Track whether AI-enabled processes are being used consistently and delivering value, not just whether tools have been deployed

AGN AI Masterclass Contributors and Judges

Executive Summary

Artificial intelligence has moved from an emerging trend to a core capability for accounting firms. For mid-sized firms, AI is becoming essential for improving efficiency, attracting and retaining talent, enhancing client service, and building resilience in an increasingly competitive market.

Delivered over two sessions, the AGN AI & Automation Masterclass focused on execution rather than awareness. Instead of showcasing tools or vendors, the programme challenged participants to answer a key question: how do accounting firms turn AI interest into AI capability?

Combining strategic insight, practical frameworks, peer learning, and expert feedback, the Masterclass culminated in participants designing and presenting AI-enabled process improvements from their own firms to an international panel of technology, consulting, and accounting experts.

This paper draws insight from three key sources:

1. The strategic framing and capability discussions from Part 1 in Amsterdam
2. The practical proposals and live scrutiny from Part 2 in Munich
3. The completed judges’ scorecards and qualitative feedback

Together, these sources offer a rare, evidence-based perspective on how mid-sized accounting firms are approaching AI adoption: identifying successes, exposing overconfidence, and highlighting gaps in leadership attention.

Three clear leadership truths have emerged:

1. Most firms are choosing sensible AI use cases, but underestimating execution complexity.
2. Technology ambition routinely outpaces change leadership, data readiness, and governance.
3. The real constraint on AI progress is not tools, but ownership — particularly at partner level.

For firm leaders, the message is unmistakable: AI adoption is no longer about experimentation, but about building repeatable capability.

Why AI Adoption Has Become a Leadership Issue

Mid-sized accounting firms face ongoing pressure from talent shortages, rising client expectations, and tighter margins scrutinised by both clients and potential investors. The market is further disrupted by consolidation, with private equity-backed platforms reshaping competitive dynamics.

In this context, AI is frequently seen as a solution to multiple challenges: improving productivity, quality, scalability, and differentiation. Yet, despite the enthusiasm, many firms remain trapped between interest and actual execution; pilots are launched and tools tested, but firmwide impact is rare.

At the Amsterdam Masterclass, a prominent theme was the widening gap between AI awareness and AI capability. While many partners understand AI’s importance, few can point to embedded, governed, scalable AI-enabled processes in their firms. This gap is not primarily technological. The tools available are more powerful, affordable, and accessible than ever. The real constraints lie in operating models, decision rights, data discipline, and change leadership. The Masterclass explicitly surfaced these constraints, making them central to the conversation.

2. Designing for Reality: The Masterclass Approach

The AI & Automation Masterclass was carefully structured to address both strategy and execution.

3. What Firms Chose to Automate — and What This Tells Us

Across the submissions, a consensus emerged around the operational areas where firms believe AI creates the most value. Rather than speculative or peripheral projects, proposals targeted core operational friction points.

Client Onboarding and Lifecycle Management

Several teams prioritised the client onboarding process, a critical stage impacting speed-to-revenue, compliance risk, and first impressions. Typical solutions included:

• AI-supported document intake and validation
• Automated KYC/AML checks with human review
• Workflow visibility across teams

Judges agreed these use cases offer clear value, but emphasised risks around ownership and data integration due to onboarding’s involvement of multiple systems and stakeholders.

Audit and Assurance Workflow Intelligence

Audit-focused submissions aimed to reduce coordination friction, not replace professional judgement.

Common features included:

• Centralising client touchpoints in CRM
• Automating reminders, agendas, and meeting documentation
• Using AI to surface risks discussed in meetings

These proposals received strong scores for client experience and efficiency, with judges highlighting the need for robust audit trails and clear human accountability.

Core Accounting and Compliance Automation

Other teams addressed compilation engagements, VAT reviews, and transaction processing, aiming to:

• Reduce manual data handling
• Improve consistency and timeliness
• Free staff capacity for higher-value work

Judges recognised the maturity of these ideas but raised concerns about data quality, exception handling, and cross-department coordination.

Internal Operations and Enablement

Some proposals focused on foundational automations like inbound mail handling and document management. While less visible to clients, these initiatives are crucial for scalability and were viewed as sensible starting points for firms early in their AI journey.

Winners of the Masterclass

The quality of the final presentations reflected the creativity and practical thinking demonstrated throughout the Masterclass. Congratulations to the 2026 NextGen AI & Automation Masterclass winners.

First place was awarded to Megan Healey, Audit & Accounts Manager at Ballards (UK). Megan’s project proposed using AI to automate incoming mail and create a scalable knowledge bank to support wider AI adoption across the firm. The judges praised the project’s commercial focus, practical implementation and strong client service rationale.

Second place went to Felix Galkowski at MTG (Germany). His automation solution transformed a 10-day workflow into a streamlined, technology-enabled process. His project demonstrated strong technical capability and highlighted the potential for firmwide automation.

Both projects exemplified the practical, business-focused innovation the Masterclass was designed to support and demonstrated how NextGen professionals are helping shape the future of their firms.

4. What the Judges’ Scores Really Reveal

Despite varying individual scores, clear patterns emerged in the completed score sheets.

Strengths: Problem Clarity and Ambition

Most proposals excelled in:

• Clearly articulating the business problem
• Logically linking process pain to the proposed solution
• Aligning ambition with real firm challenges

This indicates that firms have largely moved beyond random AI experimentation, selecting credible and relevant use cases.

Weaknesses: Execution Readiness

Scores dipped in areas related to execution, such as:

• Data readiness and integration
• Change management planning
• Ongoing governance and ownership

Judges often commented that while the ideas were strong, many lacked a convincing path to firmwide adoption.

The Partner Gap

The most consistent theme was the absence of partners in execution. While sponsors were mentioned, judges questioned whether partners would genuinely change behaviour, dedicate time, and enforce adoption beyond the pilot stage. This points to a leadership rather than a technology issue.

5. Why Good AI Ideas Struggle in Practice

Despite varying individual scores, clear patterns emerged in the completed score sheets.

6. What Firm Leaders Should do Differently

What leaders should do differently	What this means in practice	Upside gain if done well	Risk if ignored or done poorly
Choose fewer, better AI initiatives	Prioritise a small number of high-impact processes that clearly affect clients, risk, or staff capacity, rather than running multiple disconnected pilots.	Faster time to value, clearer returns on investment, and organisational focus on what really matters.	Fragmented experimentation, pilot fatigue, and limited firmwide impact despite significant effort.
Assign real accountability	Appoint a named business owner for each AI-enabled process with authority to make decisions, allocate time, and enforce adoption.	Clear ownership accelerates execution, improves adoption, and reduces internal friction.	Initiatives stall between departments, accountability is diluted, and AI becomes perceived as ‘someone else’s problem’.
Invest in data readiness early	Treat data structure, integration, and quality as foundational work that must be addressed before scaling AI solutions.	More reliable AI outputs, fewer exceptions, and reduced rework as initiatives scale.	AI tools underperform, confidence erodes, and teams revert to manual workarounds.
Treat change management as core work	Plan training, communication, and behavioural change from the outset, with visible leadership involvement.	Higher adoption, faster cultural acceptance, and sustained benefits beyond the initial rollout.	Resistance builds quietly, adoption stalls, and AI initiatives fail despite technically sound solutions.
Listen to and empower NextGen leaders	Actively involve NextGen professionals who work closest to the processes being automated and have real-time insight into inefficiencies and opportunities. Pair their energy with senior sponsorship.	More relevant use cases, faster learning cycles, and stronger future leadership capability within the firm.	Valuable insight is missed, initiatives lack credibility on the ground, and NextGen engagement is reduced.

Conclusion

The question for mid-sized accounting firms is not whether AI will matter; it already does. The real challenge is determining who will lead its integration into the firm’s operating model.

The Masterclass demonstrates that capability exists within firms, especially among NextGen leaders. Success depends on whether leadership converts that capability into intent, ownership, and sustained action.

Ultimately, AI is not a technology programme but a test of leadership.

---

Contact:

For further information on this topic or anything relating to the AGN International Association of Accounting and Advisory Firms or to become an AGN member, please email your closest AGN Regional Director (see below) or go directly to www.agn.org.

---

Copyright © 2026 AGN International Ltd. All rights reserved. No part of this publication may be reproduced, distributed, or transmitted by non-members without prior permission of AGN International Ltd.