A new study by the Massachusetts Institute of Technology warns that the corporate rush into generative artificial intelligence is yielding scant financial benefits. The report, “The GenAI Divide: State of AI in Business 2025,” finds that 95 % of the 300 pilot projects it examined produced no measurable impact on profit and loss, despite an estimated $30 billion to $40 billion that U.S. companies have ploughed into the technology. Only one in twenty pilots generated “rapid revenue acceleration,” with researchers blaming brittle workflows, misaligned use-cases and a misallocation of budgets to sales and marketing instead of back-office automation. The sobering data landed as OpenAI chief executive officer Sam Altman publicly cautioned that investors are “overexcited” about AI and that “someone will lose a phenomenal amount of money.” Altman nonetheless continues to court record sums, seeking a secondary share sale that would value OpenAI at roughly $500 billion and telling reporters the company may spend “trillions” of dollars on data-centre construction. The juxtaposition of bubble warnings and ever-larger fundraising plans has intensified debate over whether AI valuations can be sustained. Market jitters are already evident. Shares of software stalwarts Salesforce, Adobe and ServiceNow have fallen at least 16 % this year, wiping about $160 billion from their combined capitalisation, while credit investors pour fresh billions into AI-linked infrastructure. Some analysts, including Wedbush’s Dan Ives, contend the industry is still in its early innings, but the MIT findings and Altman’s remarks have sharpened comparisons with the late-1990s dot-com boom and raised the stakes for the sector’s next set of earnings reports.