AI Use Statement

Summary

We use AI to work faster, not to replace judgment. A person reviews everything before it is published or sent, and a person is accountable for everything we publish.

Where we use AI

We use Anthropic's Claude API to help with:

• Research and drafting: summarizing public sources, suggesting outlines, drafting first versions of articles and directory descriptions, and editing for clarity.
• Directory upkeep: drafting and updating Company and Product descriptions from public information.
• Inbox triage: summarizing and sorting incoming messages, and drafting replies.
• Audience statistics: classifying contacts by role or seniority and organization type (for example, "in-house, GC/CLO") so we can report aggregate audience composition.

The AI chat assistant on our home page is provided through Algolia and uses models from Anthropic (Claude) and Google (Gemini) to answer questions from our directory. Its answers are generated in real time and are not reviewed by a person before you see them, so they can be incomplete or wrong. Please confirm important details with the Company before relying on them.

Human review

• A person reviews, edits and approves every article, listing description and newsletter issue before it is published.
• AI never sends an email on our behalf. Draft replies are reviewed and sent by a person.
• We check facts, figures, quotes and citations against original sources. AI output is never treated as a source.
• We do not publish AI-generated images of real people.

What data goes to AI providers

• For content work, we send public information and our own drafts.
• For inbox triage and contact classification, we send the relevant message or contact details, such as name, job title, company and message text.
• For the AI chat assistant, the questions you type go to Algolia and to Anthropic and Google to generate answers. Please do not enter confidential, privileged or sensitive personal information.
• We use Anthropic's commercial API. Under Anthropic's commercial terms, API inputs and outputs are not used to train its models by default.
• We do not use AI to make decisions that have legal or similarly significant effects on anyone.

See our Privacy Policy for more on how we handle personal information.

Contributors

Guest authors must disclose significant AI use and verify their own facts under our Creator Agreement.

Mistakes

AI makes mistakes, and so do people. If you spot an error, tell us under our Corrections Policy.

Contact

Questions about how we use AI: hello@corporatelegal.tech