AI Will Continue to Grow in 2026 — But It Will Face Major Challenges Along the Way

Artificial intelligence made large and surprising advances through 2024 and into 2025. People began conversing with AI ‘digital resurrections’ of deceased loved ones, using AI-powered toothbrushes, and engaging with AI-driven recommendation systems embedded in nearly every app. OpenAI was valued at US$150 billion and claimed progress toward AI systems more capable than humans. Google’s DeepMind made similar assertions. But researchers warn the path forward is not smooth. Training data has hit a wall as high-quality sources are being exhausted. Scaling laws — the principle that bigger models are smarter models — appear to be plateauing. New approaches such as reasoning-based models attempt to overcome this, but at higher computational cost. Experts argue that regulation, global consensus on safety, and data ownership frameworks must catch up with deployment speed before the technology’s 2026 trajectory becomes ungovernable.