EY Centralizes AI Governance Amid Rising Costs

The rise in costs associated with AI models is prompting organizations to strengthen their governance. EY has implemented an "invisible" router that has reduced its token consumption by 60% since April and has published a survey where 98% of decision-makers report reevaluating their approach. In this context, the firm has created an AI Value Realization Office, which is expected to become fully operational in the coming months.

Costs: 60% Savings at EY and 98% of Decision-Makers on Alert

The shift to new pricing models among AI providers is increasing the pressure to use these tools effectively and control spending. EY has deployed an "invisible" AI router behind some of its specialized tools, guiding employees to the most suitable model for each task. According to EY, this device has reduced token consumption by 60% since its deployment in April, alongside other governance measures. From the clients' perspective, Dan Diasio notes that cost-saving levers are gaining importance: many had experimented when certain expenses were subsidized by providers and are now facing the true cost, seeking EY's guidance on how to scale their organization. An EY US AI Pulse survey, published in July and conducted between April and May with 534 senior decision-makers, indicates that 98% of respondents say that token spending has led them to reconsider their approach.

EY Establishes an Office to Govern Spending, Usage, and Returns

EY has created an AI Value Realization Office to steer its strategy and centralize its spending, aiming to measure its impact. According to Dan Diasio, this office is expected to be fully operational in a few months. Its roadmap includes governance of AI spending, seeking measurable returns, monitoring usage, prioritizing initiatives to develop, and overseeing the effects of AI on jobs. This office will also handle the value strategy by assessing, beyond dollar-for-dollar calculations, whether projects are changing the company's performance. For Diasio, this body will help direct budgets toward the most significant opportunities and is a "critical part" of clients' AI journeys.

The Value of AI Crosses Silos, According to EY-Parthenon

Budgets are most often managed by department, while, according to Dan Diasio, AI-related spending and potential returns are spread across IT, finance, sales, HR, and operations. Funded by function, projects end up addressing local use cases and often leave value "on the table," he emphasizes. Research from EY-Parthenon quantifies this issue: 75% of the potential value of AI would come from horizontal flows, compared to 25% for projects confined to a single function.

Internal Showcase and Investments: EY Mentions $1.4 Billion for EY.ai

The Big Four present themselves as "client zero" of AI by testing these technologies on their own teams and processes to show clients what works and sell them similar approaches. EY reports having invested $1.4 billion in 2023 to build the foundations of its EY.ai platform. At the same time, the creation of a dedicated corporate function recalls the emergence of HR during the Great Depression and the formation of treasury functions in the 1970s in response to currency risk: the AI Value Realization Office consolidates responsibilities that were previously dispersed among finance, technology, and other teams. Dan Diasio notes that most clients have not yet established dedicated offices, while envisioning that this will become a function over time. So far, many have funded AI by tightening budgets and reallocating funds. Looking ahead, he believes that placing more emphasis on structured AI governance, through a function like this office, will be one of the determining factors, with attention shifting toward the target organization.