Inspect observed AI provider disclosures by industry and company size. Distinguish the current coverage snapshot from a historical adoption curve.
Reviewed coverage sample · source reviews to (UTC)
Real public-source records, manually reviewed. This is a dated snapshot, not a continuous live feed or a representative market survey.
Dated AI adoption evidence
Source reviews 2026-09-10 to 2026-10-15 · 3 distinct companies in the selected model-provider sample.
Each company counts once per provider. Shares can total more than 100% because companies can disclose multiple providers.
Time coverage
Observation period
Covered companies
Comparability
2026-09-10 to 2026-10-15
3
Assembled sample with individual review dates
Earlier months
Unavailable
No comparable company panel
This assembled sample does not establish a monthly trend. Company-size values are not provided by these model-provider disclosures.
Delivery route: not established
The records name providers, but do not establish direct API use versus Azure OpenAI, Bedrock, or Vertex. All 8 records remain unclassified by delivery route; no route split is inferred.
Public-source observations. Observation dates are review dates, not adoption dates.
Company
Vendor / signal
Industry / scope
Observed (UTC)
Evidence
Intercomintercom.com
OpenAIPresence observed
Customer service softwareDefault US-hosted service: AI processing
The company’s published subprocessor disclosure names this provider for the stated purpose.
What this adoption curve currently contains
The reviewed provider sample begins in September 2026. Selected records have since been rechecked; these individual source reviews do not create a comparable monthly series or establish a trend. The earlier Intercom disclosure is a source-specific historical comparison, not a representative prior-year company panel.
Enterprise AI adoption in 2025
We do not have a comparable 2025 observation cohort. This page does not reconstruct earlier enterprise adoption from current lists. A genuine year-over-year comparison requires stable entity mapping and consistent historical source coverage.
Generative and agentic AI adoption
A provider listed for AI processing may support several features and models. That label alone does not establish that an enterprise deployed an agentic workflow, which users have access, or what proportion of workloads use generative AI.
An evidence-led adoption framework
Separate provider availability, disclosed processing, confirmed deployment scope, and measured usage. Record changes at the level the source supports. Data residency, permissions, evaluation, and integration can affect adoption, but a disclosure alone cannot quantify those challenges.
A public job board supports an ATS observation. A subprocessor disclosure identifies a provider that may process data for particular services. A vendor customer story describes a published relationship. These sources have different scopes and are labeled separately.
Counts cover the records on these pages, not the whole market. No observation in our covered sources does not mean a company does not use a vendor. Source recency and unresolved conflicts affect confidence; observations do not establish spend, renewal timing, or customer churn.