Thought Leadership

1. From Automation to Autonomous AI Agents

While early technology centered on static Robotic Process Automation (RPA), firms across APAC are evolving toward agentic AI workflows.

In Australia, Hall Chadwick adopted a bottom-up strategy to eliminate repetitive work. By empowering junior staff to identify operational bottlenecks, they built internal tools like "Baxter"—using Copilot and SharePoint—to automate routine Goods and Services Tax (GST) preparation.

In India, Jigar Doshi highlighted the shift from fixed RPA bots to dynamic AI Agents. Unlike traditional automation that breaks when government portals update, AI Agents independently monitor tax portals, summarize notices, and draft responses. In Vietnam and across the region, Vlad Savin shared that Acclime is developing an overarching digital platform connecting disparate software tools into a central hub backed by an agentic workflow.

2. Enhancing Audit Quality and Capital Market Compliance

In audit and assurance, mid-sized firms are demonstrating that practical AI deployment streamlines compliance without sacrificing oversight.

Jenny Jiang explained how Shanghai CPA uses a three-part AI model for capital market client acceptance:

1. Automated Web Scraping: Gathering background data, financial histories, audit opinions, and regulatory penalties.

2. Six-Dimension Risk Scoring: Integrating an AI fraud model to generate risk scores across key audit dimensions.

3. Partner Judgment: Applying partner-level judgment to findings to determine necessary risk mitigations.

By letting AI handle data aggregation, audit teams free up valuable time to focus on strategic risk discussions and critical thinking.

3. Reimagining the Business Model: Selling Outcomes, Not Hours

As AI assumes routine groundwork, traditional hourly billing is under scrutiny. Panelists agreed that professional services are shifting away from selling raw information toward delivering high-value solutions.

As junior staff spend less time on manual entry, they can step into advisory roles earlier. This enables firms to transition toward fixed-fee structures or percentage-based outcome pricing. As Jigar Doshi noted, accountants should view themselves less like general practitioners charging for time and more like specialized surgeons charging for experience and impact.

4. Safeguarding Quality: Mitigating Hallucinations and Data Privacy

Despite its transformational power, panelists emphasized that AI remains an assistant, not a source of truth:

• Hallucinations & Skepticism: AI models generate answers even when uncertain. To prevent reliance on inaccurate outputs, firms must preserve professional skepticism and mandate human checkpoints—especially for senior partners evaluating technical advisory work.
• Data Security & Privacy: Confidential client data must never be pasted into public LLMs. Mitigations include enterprise vendor agreements, disclosing AI usage in engagement letters, and using local scripts for audit sampling.
• Disaster Recovery: As firms move to cloud subscription models, robust contingency plans are essential to protect against outages, API changes, or vendor lock-in.

5. Training the Next Generation of Professionals

A major concern among firm leaders is ensuring junior staff develop core accounting fundamentals when routine tasks are automated.

Panelists outlined a balanced approach:

• Foundational Periods: Requiring new hires to perform manual reconciliations and testing for their first 3 to 6 months to master core principles before using AI tools.
• Prompt Engineering & Transparency: Teaching proper prompting techniques while requiring staff to document in working papers where and how AI was utilized.
• Fostering Innovation: Running internal AI competitions to encourage staff to share efficiency ideas and workflow enhancements.

When hiring, panelists prioritize core accounting expertise, critical thinking, and adaptability over pure technical AI skills.

6. Actionable Advice for Firm Leaders

For practices wondering where to start in the next 3 to 6 months, the panel offered four key recommendations:

1. Start Small: Don't wait for a massive platform. Identify 1 to 5 specific use cases to automate first.

2. Dual Strategy: Split your roadmap into internal efficiencies (in-house workflows) and external value propositions (client communication and service expansion).

3. Tailor Prompts for Strategy: Use AI tools to help analyze your firm's specific operating model and draft a tailored implementation plan.

4. Prioritize Human Relationships: Technology is an enabler, but client intimacy, trust, and personal interactions remain the true differentiators of a successful firm.

Technology will continue to accelerate, but humans will always remain at the core of the work process—making human judgment and professional experience more vital than ever in an era of rapid artificial intelligence development.