Accenture is reportedly looking for ways to control employee use of generative AI after internal adoption began creating unpredictable costs for the consulting company.
The reported restrictions follow an earlier push to make AI part of everyday work. Accenture had encouraged employees to use its internal tools regularly, with senior staff reportedly told that AI adoption could affect their eligibility for promotion.
Routine Tasks Are Consuming AI Budgets
According to reporting covered by TechCrunch, Accenture is examining how employees consume its available AI tokens. The concern is that workers may use expensive models for basic assignments, including converting PDF documents into presentation slides.
Tokens are the units AI providers use to measure the text and other information processed by their models. Longer prompts, uploaded documents, repeated revisions, and complex agent-based workflows can quickly increase the number consumed.
The report was based on leaked audio from an internal meeting involving Justice Kwak, Accenture’s agentic AI strategy lead. Kwak reportedly said AI spending was becoming material to the company’s cost structure while remaining difficult to predict.
He also said senior financial and technology executives were asking whether the company was receiving sufficient value from the money spent on AI. Accenture has not publicly confirmed the recording or detailed any company-wide token restrictions.
Earlier AI Push Linked Usage to Career Progression
The reported cost controls come only months after Accenture increased pressure on employees to adopt AI. An internal message previously reported by The Guardian said regular use of the company’s AI tools would be considered when evaluating some senior employees for leadership roles.
Accenture had trained approximately 550,000 employees in generative AI and invested heavily in integrating the technology into its consulting services. The company has also continued reporting increased demand for large-scale AI programmes among its customers.
The two positions are not necessarily contradictory. Accenture can require employees to become proficient with AI while limiting applications that cost more than the time or labour they save.
Companies Face Pressure to Measure AI Returns
The issue reflects a growing challenge for businesses that rapidly expanded access to generative AI. High adoption figures can demonstrate enthusiasm, but they do not show whether each use improves productivity, revenue, or customer outcomes.
Agentic systems can be particularly difficult to budget because they may perform several model calls while completing a single assignment. Research has also found that higher token consumption does not always produce better results.
Companies are therefore placing more attention on cost controls, model selection, and measurable returns. The next stage of workplace AI adoption may depend less on how frequently employees use the tools and more on whether each use produces enough value to justify its cost.
Featured image credits: Wikimedia Commons
For more stories like it, click the +Follow button at the top of this page to follow us.