95% of companies using generative AI getting zero return: report

SAN FRANCISCO (KRON) — In what could be the latest sign that the AI bubble, if not bursting, may be beginning to let a little air out, a new report finds a majority of companies using generative AI aren’t getting much in return. The MIT report, “The GenAI Divide State of AI In Business 2025,” found that 95% of companies using generative artificial intelligence are getting zero return.

The report, which comes on the heels of NVIDIA stock nosediving and Sam Altman saying the quiet part out loud, was based on 52 structured interviews with stakeholders, surveys of 153 leaders, and analysis of over 300 public AI deployments.

While a vast majority of GenAI deployments achieve virtually no measurable P&L (profit and loss) impact, some 5% are extracting millions in value. This divide, according to the report, doesn’t seem to be driven by the quality of the model or regulation, but determined by approach.

Widely adopted tools like ChatGPT and Copilot have been explored or piloted by 80% of organizations. But rather than enhancing P&L performance, they “primarily enhance individual productivity,” the report states.

More eyebrow-raising is that enterprise-grade systems, whether custom or vendor-sold, are being “quietly rejected,” with only 20% of organizations reaching pilot stage, and just 5% reaching production, despite 60% of organizations evaluating such tools.

“Most fail due to brittle workflows, lack of contextual learning, and misalignment with day-to-day operations,” the report states.

The problem, the report goes on to state, is not infrastructure, regulation, or talent. It’s learning. Most GenAI platforms don’t retain feedback, don’t adapt to context, or improve with time.

Of the nine major industry sectors, only two — tech and media — have shown clear signs of structural disruption due to AI, the report found.

The report also called the idea that “AI will replace most jobs in the next few years,” a myth. MIT’s research found “limited” layoffs from GenAI and only in industries that are “already affected significantly” by the technology. The report found that while AI is undoubtedly having an impact on headcounts, it was not in the form of mass layoffs.

Rather, it was manifesting through not backfilling positions as they become open and the reduction of previously outsourced roles. “Organizations that have crossed the GenAI Divide,” the report said, “demonstrate measurable external cost reduction while slightly decreasing internal headcount.”

The report concluded that organizations that successfully integrate generative AI do three things well:

• Buy rather than build
• Empower line managers over central labs
• Select tools that integrate deeply while adapting over time

“The next wave of adoption will be won not by the flashiest models, but by the systems that learn and remember and/or by systems that are custom built for a specific process,” the report said.

Also of note, that while more than half of generative AI budgets are currently devoted to sales and marketing, the MIT report found the biggest ROI was in back-office automation through eliminating outsourcing and slashing external creative and content costs.