Big Four firm KPMG announced the launch of its new multi-agent AI platform, Workbench, which will serve as a foundational and single AI platform, designed to underpin its client delivery platforms as well as its other AI platforms such as KPMG Digital Gateway (Tax), KPMG Velocity (Advisory) and KPMG Clara (Audit). The capabilities are available internally to KPMG professionals and are also available to deploy for clients.

“The next phase of AI will be defined by platforms that scale,” said Steve Chase, US vice chair of AI and digital innovation. “Workbench is KPMG’s single, global AI platform—built on an interoperable architecture that supports agent-to-agent coordination and multi-model flexibility. It’s the foundation for how we’re scaling AI across our business and for our clients —with confidence, agility, and global impact.” 

The platform sports a network of 50 AI assistants, or agents, that interact with each other across multiple sectors with nearly a thousand more currently in development. They’re meant to work as “digital teammates” alongside KPMG professionals. Working on Microsoft Azure infrastructure, Workbench is conceived of as a “one stop shop” platform that will serve up tools and agents to KPMG professionals within the systems they work in every day. Most will never directly interact with the platform itself, as the assistants will be retrieving solutions for them. To do so it uses interoperable, agent to agent communications to bring together capabilities from across the KPMG ecosystem of alliance partners (e.g. Oracle, Salesforce, ServiceNow, Workday) to best address the task at hand. 

KPMG Workbench enables team members to automate complex, multi-step processes from client onboarding to regulatory reporting. Beyond internal use, private instances of KPMG Workbench will also be available to clients from across industries to help develop and manage their own digital workforce. KPMG said that clients can maintain full control of how their data is stored and processed and manage diverse risk and governance needs, helping them to meet local and global regulatory requirements. KPMG is also certified in ISO 42001, which concerns AI Management Systems.

While the precise definition can vary depending on who is asked, very broadly agentic AI could be described as software that is capable of at least some degree of autonomy to make decisions and interact with tools outside itself in order to achieve some sort of goal—whether booking a flight, sending a bill or buying a gift—without constant human guidance. Agents are not necessarily new, but the rise of generative AI has made them much easier to make and use, as doing so no longer requires specialized coding skills. Since they exploded onto the scene, the need for platforms that can coordinate between agents has become clear. 

The news comes about a month after Deloitte announced its Global Agentic Network, a connected ecosystem of AI agents for business purposes to augment and automate client operations. This, in turn, came a few months after PwC announced its AgentOS platform, which connects AI agents with each other, regardless of platform or framework, into modular adaptive workflows integrated with enterprise systems. Finally, this came just a few days after EY touted the launch of its own EY.ai Agentic Platform. 

I first discovered accounting in my sophomore year of college and was immediately drawn to its zero-sum game: every trial balance must foot to zero, every debit matched by a credit; it was a perfect harmony of logic. But as I moved from the classroom to the conference room, I realized that behind each journal entry lies a human story far richer than any spreadsheet.

At EY, I spent years in the international tax practice, working on mergers and acquisitions, IP onshoring and cross‑border reorganizations, an environment where the answer almost always began with “it depends.” Every multistep plan we proposed carried consequences across jurisdictions, and our task was to bring every stakeholder along, weighing trade‑offs in tax savings, implementation costs and regulatory scrutiny. Gaining that buy‑in requires something more than technical expertise; it demands trust, built through empathy and open dialogue. And while we may lean on AI tools for speed and analysis, the trust we place in another person is fundamentally different and irreplaceable.

It is certain that AI will truly upend our way of working. When I started out in my career and had to understand what GILTI, BEAT and FDII meant and why I should care about them, I had a few paths to take: 1. Go ask my senior or manager and have them explain it to me; 2. Look it up on Google and get well and truly lost reading through the regs; or 3. Go read a BNA portfolio. Now with even the most basic AI large language model, I can ask for it to define the subject, give me an example and break it down for me as if I am a child.

The access you have today to information is comparable to when the internet first became available and with this brings the debate: ‘Will we need accountants in the future?” I understand why this question keeps getting brought up. Most people outside think we have a standardized workflow with pristine data for which we just log in at 9 and go home at 5 after working on our beloved spreadsheets. But ask anyone in the industry and they will tell you how they would love to have this kind of a lifestyle where each answer was clear and there was a definitive way forward. 

As AI technology evolves, so too will the tools that have shaped our careers. Every day, new startups launch with the promise of revolutionizing how we work. Before long, I believe all the manual tasks we once performed will be fully automated. I look forward to the day when I can simply extract a trial balance from the system and generate a tax return that automatically applies book‑to‑tax adjustments, tracks every business change from the year, and delivers a complete, compliant filing. I’ll shed tears of joy the first time I never have to hand‑fill a Schedule Q again.

But that’s when my real work will begin. I will review each return not only to confirm that the numbers are entered correctly, but also to ensure they make sense in context. I’ll seek to understand the story the business is telling: Why did certain figures move? What does this reveal about the company’s strategy? And based on those insights, how can we design a proactive plan for the coming year? To answer these questions, I’ll draw on every lesson from my critical thinking courses and ask why.

AI shouldn’t be viewed as a career threat but as a powerful partner in our work. After all, our clients’ finances are deeply personal, tied to their dreams, anxieties and life milestones, and emotions inevitably come into play. With regulations and tax laws shifting daily, clients will look to us not just for compliance, but for someone who can translate figures into a meaningful narrative. You personally know how many questions you got asked over the last few months asking you to predict what is going to happen this year. That’s where human judgment is irreplaceable: knowing when to dig deeper, which details to emphasize, and how to guide clients through complexity with clarity and compassion.

In the age of AI, the professionals who will excel are those who invest as much in empathy, communication and critical thinking as they do in technical skills. They’ll build workflows that require every AI‑generated insight to pass through a human lens and welcome the moments when financials reveal a story of resilience or ambition. AI will continue to accelerate change, but behind every algorithmic recommendation and every line on a tax return, there must be a person ready to listen, interpret and guide. Because at its core, finance will always be personal.

Grant Thornton Advisors is adding Grant Thornton Switzerland/Lichtenstein and Grant Thornton in the Channel Islands to the multinational platform it launched earlier this year. Both transactions are expected to close later this year.

In January, Grant Thornton Advisors combined with Grant Thornton Ireland to create an integrated international firm. In April, it announced agreements with GT firms in Luxembourg, the United Arab Emirates and the Cayman Islands, as well as GT Netherlands in May.

The firm is backed by private equity firm New Mountain Capital, which acquired its majority stake in March 2024 after selling a majority stake in Top 100 Firm Citrin Cooperman. As a result of the PE investment, Grant Thornton took on an alternative practice structure, splitting its non-attest services into Grant Thornton Advisors and its audit and assurance services into Grant Thornton LLP.

By adding firms in Switzerland, Liechtenstein and the Channel Islands, Grant Thornton is expanding its geographic footprint and increasing its total headcount to 13,5000 professionals across nearly 60 offices over the Americas, Europe and Middle East. 

“We are very pleased to have our colleagues in the Channel Islands, Switzerland and Liechtenstein join our differentiated and expanding platform,” Jim Peko, CEO of Grant Thornton Advisors, said in a statement. “We’re building the world’s most talented team — delivering seamless offerings through an expanded footprint. The result: an unparalleled client experience and unmatched quality.”

Adam Budworth, managing partner of Grant Thornton Channel Islands, said in a statement: “This is an exciting opportunity to support our growth in the Channel Islands with access to new service offerings, technologies and investment capital. Joining the platform will only enhance the reputation of the Channel Islands on a bigger stage, while at the same time creating unique opportunities for our people.”

“I am delighted about this positive development and am convinced that it is the right step for our firm in the current turbulent market environment,” Erich Bucher, CEO of Grant Thornton Switzerland/Liechtenstein, said in a statement. “It opens up completely new perspectives for us and will enable us to push ahead with our growth strategy much more quickly.”