OpenAI chief executive officer Sam Altman warned that the artificial-intelligence industry is overheating, telling The Verge and other outlets that investors are “overexcited” and that today’s enthusiasm resembles the late-1990s dot-com bubble. Altman, whose company is seeking a secondary share sale that could value it near $500 billion, said some backers are likely to be “very burnt” even though he still views AI as a transformational technology. His remarks landed as a Massachusetts Institute of Technology study highlighted how little payoff most companies are seeing from the technology. The NANDA initiative’s “GenAI Divide” report found that 95 percent of corporate generative-AI pilot projects have failed to achieve measurable profit-and-loss impact despite $30 billion to $40 billion in enterprise spending; only 5 percent of pilots delivered rapid revenue gains. The twin signals of exuberant funding and disappointing results rattled financial markets. On Aug. 19 the tech-heavy Nasdaq Composite fell 1.4 percent, its steepest one-day decline in almost three weeks, while the S&P 500 information-technology sector slid about 2.5 percent for the week through Aug. 20. Chipmaker Nvidia, which has become the emblem of the AI trade, dropped 3.5 percent in a single session and roughly 5 percent over the week; software group Palantir tumbled 9.4 percent the same day. Some strategists say the pullback is a healthy correction after a 50 percent surge in tech shares since April, but others caution that sky-high valuations leave the sector vulnerable if promised productivity gains fail to materialise. While Altman insists long-term value will be “huge,” the MIT findings underscore the gulf between AI’s potential and current corporate execution, sharpening debate over whether the boom is already showing bubble-era cracks.