Enterprise spending on artificial intelligence software crossed $100 billion on an annualised basis in Q2 2026, according to estimates from both Gartner and IDC published this week, marking a milestone that analysts said would have been considered implausible three years ago when the market stood at under $15 billion.
The figure includes spending on foundation model API access (the largest category at approximately $38 billion annualised), AI-native SaaS applications such as code generation assistants and document processing tools (approximately $29 billion), AI infrastructure and operations tooling including observability and evaluation platforms (approximately $19 billion), and enterprise deployments of open-source models including hardware, fine-tuning compute, and integration services (approximately $14 billion).
The growth trajectory is striking. The market grew approximately 140% year-over-year from Q2 2025, driven by a combination of price reductions in foundation model APIs — which have fallen over 80% per token across leading providers since early 2025 as competition intensified — and a wave of enterprises completing proof-of-concept phases and moving AI workloads into production.
The industries leading adoption are financial services, healthcare, and software development. Financial services firms are the largest per-company spenders, driven by document processing automation, risk model augmentation, and AI-assisted trading research. Healthcare organisations are the fastest growing segment, with clinical documentation automation and radiology report assistance driving rapid deployment. Software development — where AI coding assistants now contribute to a meaningful fraction of production code at early-adopting organisations — accounts for a disproportionate share of the foundation model API spending.
Despite the large headline number, enterprise AI adoption is characterised by striking unevenness. Survey data shows that fewer than 20% of enterprises account for over 80% of AI software spending. Many large organisations are still in the proof-of-concept phase, and the gap between AI leaders and laggards within industries is widening in ways that analysts expect will have competitive consequences within 18–36 months.
Cost management is becoming a significant operational focus. As AI spending grows, procurement and finance teams are pushing for greater visibility into usage and ROI. A market of AI cost management and FinOps tooling has emerged, with companies like Helicone, Braintrust, and LangSmith gaining enterprise traction on the back of organisations needing to govern sprawling AI API consumption.