Thomson Reuters bets $40M on owning its AI instead of renting from OpenAI or Anthropic
$40M over two years. Thomson Reuters is betting it can own its AI stack instead of renting from OpenAI or Anthropic—and it's building on Alibaba's Qwen.
Why it matters
A major enterprise is making a strategic capital bet to verticalize AI rather than depend on third-party frontier models. This signals a shift in how large companies are thinking about AI vendor lock-in and proprietary data leverage—and it's a model (pun intended) other enterprises will watch closely.
The key facts
6 to know$40M investment over two years
Custom model 'Thomson' built on Alibaba's Qwen foundation
Model only achieves top performance when accessing Thomson Reuters proprietary content (Westlaw)
CTO Joel Hron: strategic insight is knowing 'which intelligence you need to own'
Positioning as alternative to renting from OpenAI/Anthropic
Vertical AI strategy: integrating proprietary data + custom model
Go to the source
The Decoderthe-decoder.com
Publisher excerpt: Thomson Reuters is launching "Thomson," its own language model built on Alibaba's Qwen, at a cost of about $40 million over two years. But the benchmarks only show top marks when the model can tap into the company's own content, like Westlaw. CTO Joel Hron makes the point that what matters isn't…