EY and the $600 million question

The mountain was imaginary. The bill was real. In April 2023, Ernst & Young stopped work on Project Everest, its proposal to separate audit from consulting. The reported cost had reached $600 million, including roughly $300 million of internal work. For a network paid to help other organizations make difficult decisions, this was an awkward decision to abandon.

But Everest is a useful place to begin. An accounting network is an unusual institution. It must be close enough to a company to understand its machinery, yet sufficiently independent to challenge the numbers that machinery produces. Add a consulting business, and the same brand also offers to rebuild the machinery. There is money on both sides of that boundary. There are rules about crossing it.

01 / The boundary became the business

EY belongs to the Big Four alongside Deloitte, PwC and KPMG. Its familiar initials conceal a network of separate legal entities. Ernst & Young Global Limited, the coordinating company, does not serve clients. Member firms do. That distinction sounds like small print until a global restructuring needs everyone to agree.

Everest proposed separating the businesses, with tax capabilities divided between the resulting organizations. The attraction was understandable: a consulting business freed from audit-related restrictions could pursue more work. The difficulty lay in deciding what each side needed to remain viable. Tax expertise sat across that dividing line. The US Executive Committee decided against proceeding. The design had outrun the agreement required to deliver it.

EY’s public explanation was restrained: “we are stopping work on Project Everest.” Reports of the $600 million cost also described savings from deferred projects. Treating the entire figure as an external cash bonfire would be misleading. Still, it was a substantial investment in a future the partners could not agree to inhabit.

The design had outrun the agreement required to deliver it.

An organizational lesson from Project Everest

02 / Four doors into the same client

To see what was at stake, consider what a company actually buys. Assurance includes financial statement audits and other work that helps people assess reported information. Tax teams handle compliance, planning, controversy and cross-border obligations. Consulting covers technology, operations, cybersecurity, risk and workforce change. EY-Parthenon supplies strategy and transaction services, including due diligence, valuation and restructuring.

Those services meet different moments in a company’s life. A finance director needs an audit. A business entering another country needs tax advice. A buyer needs to understand an acquisition. A technology leader needs systems that function across departments. EY can assemble several kinds of specialist around those problems, subject to independence requirements.

The model is professional-service fees, including project work and ongoing managed services. Its competition changes with the assignment: another Big Four firm for audit or tax, Accenture for implementation, strategy specialists for boardroom questions. EY’s breadth is useful when a problem crosses those categories. It is less persuasive when the assignment is narrow and a smaller specialist can do it well.

03 / The AI story starts with untidy data

EY’s technological ambitions become more interesting when they stop sounding technological. Take IBM’s tax department. Its problem was inaccessible and incomplete data, with information scattered across systems. The work required gathering and reconciling that information before anyone could do much useful with it.

EY and IBM developed EY.ai for tax, built with IBM watsonx. Their public account describes consolidating data from 36 sources, with controls and reconciliation checkpoints. This is a concrete use for enterprise AI: make a recurring compliance process less dependent on manual assembly. The published case does not justify assuming every tax department will get the same result.

The broader EY.ai platform launched in 2023 after $1.4 billion in investment. That figure describes EY’s investment, not a funding round or a client price. NVIDIA later joined EY on an agentic platform. These alliances connect technology suppliers with the tax, finance and risk knowledge required to apply their tools inside an organization.

04 / An intern inherits the partnership

Janet Truncale became Global Chair and CEO in July 2024 after beginning her EY career as an intern. Her All in strategy emphasized working across the network. After Everest, even the name carried an unmistakable message: the next effort would depend on cooperation inside the existing organization.

There is a pleasing historical complication here. Arthur Young and Alwin Ernst never met. Their successor firms came together in 1989; the global brand became EY in 2013. An institution bearing two men’s names has always depended on relationships extending far beyond either man.

Its culture also deserves scrutiny beyond recruitment language. In 2022, the US member firm agreed to a $100 million SEC penalty over exam cheating and misleading submissions to the regulator. Trust is central to the service EY sells. That episode shows why buyers should examine controls and conduct alongside credentials.

05 / What to borrow, what to ask

Everest offers a practical lesson for anyone reorganizing a business: identify the people with vetoes, then agree how shared capabilities will be divided. The IBM work offers another: begin with a specific workflow, its data and its checks. A broad ambition needs a small, testable unit of progress.

For a prospective EY client, the questions are equally concrete. Which member firm signs the contract? Who leads delivery? Which services are permitted alongside an audit? What result will be measured? Without usable data, accountable owners and agreement on scope, technology and expertise have little room to help. The yellow beam gets attention. The engagement team has to earn the fee.