TechCrunch Startup News Anthropic, Google score win by nabbing OpenAI-backed Harvey as a user
May 16, 2025
A popular legal AI tool is making a strategic shift by integrating models from Anthropic and Google, moving away from its reliance on OpenAI. This transition promises to enhance its legal capabilities significantly. Additionally, a new model benchmark leaderboard is introduced, aiming to improve transparency in AI performance. It's an exciting time for AI in the legal sector!
AI Snips
Chapters
Transcript
Episode notes
Harvey's Funding and Model Adoption Story
- Harvey was among the first startups supported by OpenAI's fund but later attracted backing from Google's GV too.
- Despite GV's investment in earlier funding rounds, Harvey delayed integrating Google's models until recently.
Harvey Expands AI Model Use
- Harvey, originally backed by OpenAI, now integrates AI models from Anthropic and Google to enhance its legal AI tool.
- Using various foundation models improves task-specific legal capabilities and AI agent development.
AI Models Excel Differently
- Different AI models excel at distinct legal tasks, as Harvey's Big Law benchmark illustrates.
- Google's Gemini 2.5 Pro is strong in drafting while OpenAI's model suits pre-trial tasks better.