🤖 AI Summary
Researchers measured enterprise LLM adoption by directly observing DNS signals across 76,084 corporate domains (companies with 500+ employees) from May 2024 to October 2025 to detect company-wide rollouts of ChatGPT, Claude, Perplexity, or Mistral. This observational approach avoids survey biases (self-reporting, selection, definition ambiguity) by only counting firm-wide deployments and explicitly excludes Gemini, Microsoft Copilot, Chinese LLMs, and vertical AI tools—an important limitation. The methodology yields concrete deployment timing, vendor choice, and multi-vendor behavior rather than intentions or pilots.
Key findings: 6% of large companies have deployed an enterprise LLM; Fortune 500 adoption tripled from 22 to 67 firms (13.4%) in a year, placing the market in an early “inning” relative to mature SaaS adoption. Industry leaders are Marketing (15.3%) and Tech (13.3%); Financial Services (6.7%) outpaces Healthcare (3.4%) due to lower experimental risk and clearer ROI. Geography is uneven: Israel leads at 12.2%, the U.S. at 9.67%, EU average 4.21% with large intra-EU variance. Competitive dynamics show heavy overlap: 48.7% of Claude customers also run ChatGPT, while only 6.5% of ChatGPT customers run Claude, and many dual deployments occur within 45 days—suggesting multi-vendor strategies rather than sequential replacement. Implications: adoption is measurable and accelerating but early-stage, shaped by sector use-cases, regulatory frictions, and evolving multi-vendor procurement.
Loading comments...
login to comment
loading comments...
no comments yet