Gartner: Explainable AI Will Drive LLM Observability to 50% of GenAI Deployments by 2028

Gartner published a new prediction today: by 2028, explainable AI (XAI) will drive LLM observability investments to cover 50 percent of all generative AI deployments, up from 15 percent today.

This means half of all enterprises running GenAI in production will have dedicated systems monitoring why models respond the way they do, and whether those responses are reliable.

Gartner highlights a shift from traditional performance metrics like speed and cost to deeper quality metrics: factual accuracy, logical correctness, and sycophancy detection. Without these trust layers, GenAI initiatives will remain confined to low-risk tasks.

The global generative AI models market is expected to surpass $25 billion in 2026 and reach $75 billion by 2029. Gartner recommends mandating verifiable XAI tracing for all high-impact use cases and prioritizing observability platforms that monitor latency, drift, token usage, cost, error rates, and output quality.

An additional warning: Gartner predicts that legal claims related to AI harm will exceed 2,000 by the end of 2026.