Stanford AI Experts Say the Hype Ends in 2026, But ROI Will Get Real

Questions continue to percolate about the impact of AI on the future of business and society at large, from speculations about a bubble to predictions that it will cause massive upheaval in our everyday life. But according to Stanford’s AI experts, 2026 may be the year that the hype cools and AI takes on heavier lifting. According to Deloitte’s 2026 Tech Trends report, the rate at which AI is evolving is motivating the urgency to shift from “endless pilots to real business value.” While the telephone reached 50 million users after 50 years, and the internet after seven years, AI tool ChatGPT reached twice that in two months and now has more than 800 million weekly users. Tangible Results AI next year may be characterized by rigor and ROI, according to Julian Nyarko, a law professor and Stanford HAI associate director. He spoke specifically about AI for legal services. “Firms and courts might stop asking ‘Can it write?’ and instead start asking ‘How well, on what, and at what risk?’” Nyarko said. “I expect more standardized, domain-specific evaluations to become table stakes by tying model performance to tangible legal outcomes such as accuracy, citation integrity, privilege exposure, and turnaround time.” Another sector that could bring about new, tangible results is AI for medicine, said Curtis Langlotz, professor of radiology, medicine, and biomedical data science, a senior associate vice provost for research, and an HAI senior fellow. Soon we’ll experience the moment “when AI models are trained on massive high-quality healthcare data rivaling the scale of data used to train chatbots,” said Longlotz. “These new biomedical foundation models will boost the accuracy of medical AI systems and will enable new tools that diagnose rare and uncommon diseases for which training datasets are scarce.” James Landay, HAI co-director and professor of computer science, predicts investments will continue to pour into AI data centers, and then the excitement will dry up. “[A]t some point, you can’t tie up all the money in the world on this one thing,” Landay said. “It seems like a very speculative bubble.” Landay expects there will be no AGI in 2026. Instead, “AI sovereignty will gain huge steam this year as countries try to show their independence from the AI providers and from the United States’ political system,” he said. For example, a country may “build its own large LLM” or “run someone else’s LLM on their own GPUs” to keep their data within their country. One Expert’s Warning With the estimates that AI applications will go deeper next year, it’s not all positive. Angéle Christin, associate professor of communication and Stanford HAI senior fellow, predicts an increase in transparency about what AI will be able to do. “[A]lready there are signs that AI may not accomplish everything we hope it will,” said Christin. “There are also hints that AI, in some cases, can misdirect, deskill, and harm people. And there is data showing that the current buildout of AI comes with tremendous environmental costs.” BY AVA LEVINSON