NEW YORK — The IEEE Computer Society’s Technology Predictions for 2026 names 26 key trends. But one fact runs beneath all of them: the hardest limits on AI’s future are not processing power or data volume. They are power generation and trust.
That is the central tension in the group’s annual forecast, released June 10. IEEE sees AI shifting from a tool to infrastructure — something as foundational as electricity or the internet. The shift is already underway. AI agents are becoming standard team members in office work. Wearable AI devices are always on. Social AI reads human emotion and adjusts its tone. Robotaxis are moving toward dense urban service networks. AI-generated video, music and documents are maturing fast.
None of that scales, IEEE argues, unless two problems get solved.
The first is energy. AI systems consume enormous amounts of electricity. The hardware is catching up to the software — in-memory computing now prioritizes performance-per-watt over raw speed. But the forecast is blunt: power generation is one of two hard ceilings on AI growth. Without enough clean, cheap energy, the whole stack stalls.
The second is trust. IEEE uses the word broadly. It includes data provenance — knowing where training data came from, whether it is reliable, whether it has been tampered with. It includes identity — verifying that an AI agent is who it claims to be, that a piece of content was generated by a known system. Without solving those, IEEE says, the infrastructure cannot hold.
Privacy questions are already surfacing around always-on wearable AI devices. Social AI that reads emotion raises obvious concerns about manipulation. AI-generated media — video, music, documents — is now mature enough to make deepfakes routine. The forecast treats these not as separate worries but as symptoms of a single underlying problem: trust cannot be assumed. It has to be engineered.
Quantum-safe cryptography is on the horizon as part of that engineering. But IEEE does not pretend the solutions are here yet. The forecast identifies the limits honestly.
Among the 26 trends, IEEE singles out medicine and engineered therapeutics as carrying the largest potential impact. AI-driven scientific discovery is listed as a key trend. AI-enabled digital twins — virtual replicas of physical systems — are becoming practical tools. Embodied AI — robots, drones, autonomous systems — is expected to scale across manufacturing, logistics and cities. The hardware is finally catching up to the software.
But all of that depends on energy and trust. Competitive advantage, the forecast argues, will move from headcount to how effectively organizations apply intelligence. The question is no longer whether AI will arrive. It is who will control energy, data and trust in an AI-driven world.
The IEEE Computer Society’s predictions are not a list of distant possibilities. They are trends the group expects to define the year ahead. And the two hardest bottlenecks are not technical in the usual sense. They are physical and social. Power generation is an infrastructure problem. Trust is a human one.
IEEE does not predict which organizations will solve them. It only states that without solutions, the infrastructure cannot hold.
