What this radar is
The Enterprise AI Radar is a monthly, opinionated briefing for CIOs and CDOs in organisations of 500 or more employees. It maps 44 AI categories across five strategic blocks and assigns each one an editorial verdict based on where the market actually is, not where vendors say it is.
It is designed to answer five practical questions:
- Should we invest in this category now, or wait for the market to mature?
- Which AI categories require board-level governance and risk attention?
- Where is commercial adoption running ahead of (or behind) technology readiness?
- What should we pilot in the next six months to stay ahead of peers?
- Which categories are becoming table stakes, and which are still differentiators?
Five strategic blocks
The radar organises 44 categories into five blocks, each representing a distinct layer of the enterprise AI architecture.
AI Foundation covers the core model and runtime layer: foundation models, LLM providers, model hubs, inference providers, agent platforms, and fine-tuning platforms.
AI Data Layer covers the data infrastructure required to feed and contextualise AI systems: data lakes, ETL pipelines, vector databases, RAG infrastructure, knowledge bases, data quality, catalog and lineage, and feature stores.
AI Enablement covers the engineering and operations layer: APIs, middleware, workflow automation, MLOps, LLMOps, CI/CD, testing and evaluation, deployment management, prompt management, and orchestration.
AI Trust and Control covers governance, risk, and compliance infrastructure: AI security, guardrails, PII protection, access control, policy enforcement, compliance, governance, auditability, monitoring and observability, traceability, red-teaming, and model risk management.
AI Value Domains covers business application areas where AI creates measurable enterprise value: marketing and sales, customer service, operations and supply chain, finance and risk, HR and workforce, product and R&D, legal and compliance, and digital workplace.
Five scoring dimensions
Each category is scored across five dimensions. All scores are in the 0 to 100% range. The overall maturity level shown in each category card is a weighted blend of these five dimensions.
Market Maturity (25%) measures how widely the category has been adopted across enterprises of 500 or more employees. It ranges from horizon signal to standard infrastructure, based on vendor density, funding activity, analyst coverage, and observable buying patterns.
Technology Readiness (20%) measures how production-grade the underlying technology is. It considers API stability, SLA guarantees, enterprise integration patterns, and the operational controls available to production teams.
Adoption Momentum (25%) measures the rate of change in enterprise uptake over the past six months, derived from news signal volume, source quality, and observable deployment patterns.
Strategic Relevance (20%) measures how central this category is to AI strategy for a typical CIO or CDO. It is weighted by frequency across analyst coverage, enterprise architecture discourse, and executive priority surveys.
Control Criticality (10%) measures how much governance, risk, and compliance attention this category demands. Higher scores warrant closer board-level scrutiny and earlier governance investment.
How ring positions are assigned
The overall maturity level is computed from the five dimensions above. Ring positions follow fixed thresholds:
Adopt (65% or above). Production-ready. Procure, deploy, and standardise with confidence. These categories have proven enterprise value, stable tooling, and broad adoption. The strategic question is no longer whether to invest — it is how to govern and scale.
Trial (45% to 64%). Worth pursuing with discipline. Run structured pilots with clear success criteria and defined exit conditions. Technology is maturing and enterprise patterns are forming, but standardising too early creates architectural debt.
Assess (25% to 44%). Early and fragmented. Build internal knowledge without committing to a vendor or architecture. The category is real but not yet enterprise-grade. Monitor market consolidation closely.
Hold (10% to 24%). Proceed with caution. Existing implementations should be reviewed; new investment should wait for market or technology consolidation. Regulatory risk or architectural instability may be a factor.
Watch (below 10%). Too nascent for investment. Track quarterly as a horizon signal. No enterprise action required today.
Signal processing
Each edition is built from 500 or more signals collected from industry sources: research institutions, regulatory bodies, analyst reports, vendor announcements, earnings calls, open-source repositories, and practitioner communities.
Signals are weighted by source quality, freshness, and relevance to the enterprise context. Each signal contributes a directional adjustment to one or more scoring dimensions. The editorial team reviews aggregated scores before publication and applies calibration where signal data is sparse or concentrated in a single source type.
Scores are directional benchmarks, not exact measurements. They reflect the balance of observable evidence at the time of publication and are updated monthly as new evidence accumulates.
A note on Hold and Watch
The absence of categories at Hold or Watch in the Q1 2026 baseline reflects the curation logic of the radar. The 44 categories were selected because they are already relevant to enterprise AI architecture. Categories that exist but have no credible enterprise application are excluded from scope rather than placed at Watch. Hold and Watch will appear in future editions when evidence supports downgrading a previously active category.