In boardrooms around the world, a specific anxiety is taking hold. It’s not just about the fear of robots taking jobs; it’s the paralyzing worry of “buyer’s remorse” in a market that moves at breakneck speed. According to Bill Briggs, Chief Technology Officer of DeloitteAs we move from AI experimentation to large-scale impact/value, this fear leads to an unbalanced investment strategy where companies spend 93% of their AI budget on the technology and only 7% on the people expected to use it.
Briggs pointed to that “93-7” split as something that really surprised him – and as a critical error. Organizations obsess over the “ingredients” (the models, chips, and software) while ignoring the “recipe,” which includes the culture, workflow, and training required to make the technology work. Briggs likened this technological approach to “trying to get paella” but ending up with “just cilantro.”
Fortune chatted with Briggs at Deloitte’s New York office, braving the holiday shopping crowds at 30 Rockefeller Center, to discuss the company’s business. 17th edition Technology trends reportan initiative Briggs has been involved with for nearly two decades. At the time, Briggs recalls, he was a senior executive, hired directly from Notre Dame as part of a larger effort to bring a technological edge to what was then primarily a tax and auditing firm. “The technology was a glimmer of an idea of what we could do more of in the future,” he recalls. THE Technology trends This report was released while consulting with companies that were establishing CTO organizations, and Deloitte did not have one. “So I came back to our CEO and said, ‘Hey, we need this whether I play the role or not. »
Briggs, who is based near Kansas City but often travels back and forth to New York and around the country, said that 93-7 split really surprised him. “I felt it during my travels, but I hadn’t been able to quantify it,” he said of the ratio. He likened this to every technology wave, where the easiest thing to do is apply the new technology to the way a business has always operated. “This incrementalism is a trap from which it is difficult to escape. »
Although Briggs did not elaborate on whether companies are spending too much or too little on AI, he said he sees too much “institutional inertia” prevailing as companies try to integrate AI into pre-existing workflows as if it were just another element, rather than reimagining their processes holistically. He recalled the famous quote from computer legend Grace Hopper that the most damaging phrase of all is “We’ve always done it this way.” To succeed in this technological revolution, he argued, leaders will have to advance what is comfortable, and this 93-7 ratio shows far too much reliance on the same old ways when the moment calls for something new.
Briggs’ comments align with massive global survey by consultancy firm Protivitireleased the same week as his new Technology trends report. During a briefing with journalists, Protiviti Fran Maxwellwho leads the global HR consulting function, said simply that “HR functions and organizations are going to have to rethink tasks. It’s not necessarily a muscle that most functions have.” And, unintentionally echoing Briggs and Hopper, he added: “You can’t solve today’s talent problems with yesterday’s talent.” »
The consequence: loss of trust and rise of ghost AI‘
To correct the 93-7 imbalance, Briggs suggested a step change in how companies view AI agents. As organizations move from “carbon-based” to “silicon-based” employees (meaning a shift from humans to semiconductor chips or robots), they must establish the equivalent of an HR process for agents, robots and advanced AI, as well as complex questions about accountability and performance management. This is going to be difficult. He discussed the hypothesis of a human creating an agent, and that agent creating five additional generations of agents. If wrongdoing occurs from the fifth generation onwards, whose fault is it? “What is disciplinary action? You’re going to put your line robot…on pause and force it to complete 10 hours of mandatory compliance training?”
The consequences of ignoring the human side of the equation are already visible in the job market. According to Deloitte Trusted ID reportreleased in the third quarter, despite growing access to generative AI in the workplace, its overall usage has actually declined by 15%. Additionally, a “shadow AI” problem is emerging: 43% of workers with access to generational AI admit to non-compliance, circumventing employer policies to use unapproved tools. This corresponds to previous Fortune report on the scourge of shadow AI, as surveys show that workers at up to 90% of companies use AI tools while hiding their use from their IT departments.
Workers say these unauthorized tools are “easier to access” and “better and more accurate” than approved enterprise solutions. This disconnect has led to a collapse in trust, with employee confidence in the AI generation declining by 38% between May and July 2025. The data supports this need for a human-centered approach. Workers who received hands-on AI training and workshops reported having 144% higher confidence in their employer’s AI than those who did not.
The fear of ‘buyer’s remorse‘
For CEOs and boards, the reluctance to address cultural change comes from a deeper fear that today’s investments will be obsolete next week. Briggs noted that executives are terrified of committing to a vendor only to face “buyer’s remorse” when a better model comes out a few days later. “CEOs and boards are afraid because they don’t want to make a commitment at the wrong time,” he said. It’s easy for them to delay committing to an AI tool because there might be another release next week, or the week after that.
Briggs likened this mentality to trying to time the stock market perfectly, but he argued that this hesitation is “almost like a pre-snap penalty” in sports. He insisted that the quickest path to progress is simply to find a solution, no matter how crowded the market may be.
The urgency to address this ratio is compounded by the arrival of “physical AI,” which extends beyond text generation to robotics and drones. Real-world applications already prove the value of good integration; for example, HPE saw 50% faster reporting from data to decision after deploying Zora AI.
For Briggs, the message to leaders is clear: The technology is ready, but unless leaders focus on people and cultural transformation, they risk ending up with expensive technology that no one trusts enough to use. As Briggs warned: “No matter how much traffic there is, the sooner you leave, the sooner you can get there.” »