Accenture is attempting to curb non-technical staff from exhausting corporate AI budgets on minor tasks such as converting PDFs to presentation slides. This conclusion comes from leaked audio obtained by 404 Media, which indicates the firm is witnessing a dramatic rise in token expenditure across the industry.
The situation marks a distinct change for technology firms and other organisations deploying AI. The era of unrestricted expansion has ended. Providers including GitHub have moved to charge clients per token rather than a flat subscription fee, resulting in rapid consumption of allowances. Uber recently restricted employee access to tools like Claude Code and Cursor. This followed a period where the company encouraged maximum usage, only for its CTO to reveal the entire AI budget had been depleted within four months. Accenture itself reportedly began mandating that senior staff utilise AI or face losing promotion opportunities.
The data also challenges the idea that highly skilled engineers generating vast amounts of code are solely responsible for the surge. In many instances, non-technical personnel are burning through tokens on non-specialised work.
Internal data reveals the source
Justice Kwak, Accenture’s agentic AI strategy lead, confirmed this during a recent internal meeting. According to the audio, he stated that internal data showed engineers were not driving the consumption. Instead, it was a significant number of non-engineers performing specific behaviours.
At one point in the meeting, Kwak and Eduardo Salamanca de Diego, senior manager of product management at the company’s Center for Advanced AI, began presenting on a concept described as “token ops.”
Kwak mentioned he possessed some slides despite noting people rarely use them anymore. As he prepared to present, Stuart Henderson, Accenture’s client group lead, interrupted. He joked he hoped Kwak had not just converted a PDF into images and then into markdown files. Henderson added, “I’m learning that’s one of the big token chewers. Turning PDFs into markdown: is that right?”
Kwak responded that the firm’s own data confirmed this.
“What we’re seeing right now is just rapid escalation in AI token spend,” he said. “As companies start to scale AI, moving from like simple chatbots into use cases that feature agentic workflows and automation and then enterprise-wide deployment of some of these tools like Copilot, Claude Code, and Codex, we’re hitting this inflection point where AI is becoming material to the cost structure; spend is becoming very unpredictable; and leadership, especially at the CFO, COO, and CIO level, are still asking the question of whether they’re getting value from what we’re spending on in the context of AI.”
“It’s really not a niche problem. It is a problem that every enterprise will face if they are bullish on AI, if they haven’t already,” he added. The volume of token spending is increasing exponentially as more people start using AI.
Kwak noted that after Accenture pushed for rapid AI adoption, the technology reached scale in most areas within the firm and its clients. However, this scale presented a new opportunity regarding token economics. While the total bill for AI spend is visible, attributing that spend at the token level to the specific value outcomes of projects remains unclear.
Finally, the controls designed to prevent excessive spending, such as budgeting or tiered limits, are arriving too late.
Following the Financial Times’ reporting on Accenture’s policy to force AI adoption or risk missing promotions, a spokesperson told CNBC: “Our strategy is to be the reinvention partner of choice for our clients and to be the most client-focused, AI-enabled, great place to work. That requires the adoption of the latest tools and technologies to serve our clients most effectively.”
Kwak stated Accenture plans to formally launch a product called “Token IQ” soon. The company did not respond to a request for comment.
As previously reported by 404 Media, some startups have boasted about spending more on AI than on human employees. Walmart also capped its staff’s use of AI tools following high demand.
What it means
Organisations are moving from a phase of experimentation to one of strict cost management. The ability to spend freely on AI generation is ending, forcing companies to track exactly who uses the tools and for what purpose. Simple administrative tasks are now identified as major financial drains, requiring new internal processes to monitor and limit usage before budgets are exhausted.
