This year will be the most difficult one yet for artificial intelligence, according to Deutsche Bank Research Institute. AI adoption has expanded in recent years, but Wall Street is betting that 2026 will see a reckoning of the technology — and the web of trades around the trend — as the market demands tangible returns from it. Software stocks have been experiencing some turbulence as of late as investors worry that the group may face a threat from AI. Big-name AI plays also saw sharp declines in Tuesday’s market rout after President Donald Trump intensified his push for U.S. control of Greenland. The S & P 500 tech sector slid more than 2%, while AI chip darling Nvidia lost nearly 4% and Google parent Alphabet fell 2%. Broadcom shed almost 5%. NVDA 1Y mountain Nvidia in the past 12 months This year will likely be a tough one for AI as three key themes emerge, according to a Tuesday note co-authored by Adrian Cox, analyst for Deutsche Bank Research Institute. They are disillusionment, dislocation and distrust. First, AI will undergo a period of disillusionment as benefits tied to the technology remain more visible to Silicon Valley and savvy early adopters, as opposed to executives who want to see a sharp pickup in revenue generation, Cox said. Talk of rollouts of agentic AI models often skips over the complication of integrating them into company workflows, he said. “Generative AI will be transformative but not right now,” Cox wrote. “As pilots move into production, enterprise users are confronting inherent limitations such as accuracy; the difficulty of applying it in unpredictable real life; and the fact that in many areas it will be a long time, if ever, before it is more economical than human labor.” Second, this year will also bring rising dislocation to the AI space, according to the analyst. He referred to a widening gap between demand and capacity due to bottlenecks, energy grid constraints and talent shortages. Private AI players ChatGPT maker OpenAI, Anthropic and xAI could feel pressure as they shore up funding and compete against their hyperscaler rivals, according to the analyst. OpenAI could face the most scrutiny, Cox said, particularly after Apple earlier this month chose Google’s Gemini model to power its AI features. “This year may be make or break for standalone AI model makers,” the analyst wrote. “OpenAI is particularly extended and may be most at risk as it seems not yet to have found a workable business model to cover its reported cash burn of $9bn last year and likely $17bn this year.” He added that while Google and other rivals are producing comparable models financed by in-house data centers, OpenAI’s moat is “relatively shallow and its path to success appears to be looking narrower and narrower.” Last, Cox anticipates AI will face increasing distrust this year. This will be reflected in anxiety around AI-driven job displacement, copyright and privacy lawsuits, data center investment impacts on power and water supplies and geopolitical competition, the analyst said. Fears of an AI race are growing between the U.S. and China as the technology becomes a tool for countries seeking self-sufficiency, he noted. “Anxiety about AI will go from a low hum to a loud roar this year,” Cox wrote.