software AM market analysis — 2026-09-09

Software’s twin engines, subscription predictability and AI-native monetisation, are converging rather than competing this session. Source commentary suggests that businesses controlling proprietary compute and selling model access through APIs may be best placed to capture attractive returns from the current wave of AI infrastructure spending, though that return depends heavily on execution rather than exposure alone.

Recurring-revenue software continues to prioritise forecasting accuracy, billing flexibility and usage-linked pricing, with vendors still migrating away from legacy licensing towards subscription and consumption models. GitLab’s latest commentary frames AI as an incremental monetisation layer sitting alongside its established seat-based business, rather than a replacement for it, while Kiro’s positioning points to agentic tools expanding beyond code completion into terminal workflows, local collaboration and deployment.

The read-through is that AI is becoming embedded in the software-development lifecycle, but the sector’s underlying economics still hinge on whether AI features convert into durable usage and pricing power, and whether compute intensity stays in line with monetisation. Packaging and workflow ownership look like the differentiators worth watching as agentic platforms compete with narrower coding assistants. This is information only, not financial advice.