top of page

the RIDE dataset:
Redline Ideation for Deal Efficiency

Thank you for your interest in advancing AI's ability to enhance legal work!

Ready to submit a redline?

What is RIDE?

What is RIDE?

RIDE will be an open-source dataset to test generative AI's ability to help human attorneys propose creative win-win solutions.

 

Would it be nice if we can leverage LLMs' creative abilities to help break negotiation deadlocks? To get there, we first need to teach LLMs what the typical deadlocks are.

 

This project is led by The Atticus Project in collaboration with AI researchers at Stanford University and ETH Zurich, and student contributors. We hope to publish the dataset and one or more research papers later this year.

What does the redline collection look like?

Contribute: Get Started

  1. FIRST REDLINE: Contributors will redline limitation of liability & indemnification clauses in a sample enterprise software license to reflect "market" position (i.e., what's fair to both parties). You can pick either Buyer or Seller by downloading the relevant file below.
    When you're done with your markup, submit it by pressing the button at the top of the page.

     

  2. REVISIONS: After you submit, you can check out a Google Drive folder of other attorneys' (anonymized) markups. Revise your prior redline based on a review of the other contributors' redlines to better reflect what is truly "market". I expect after this second round of redlines, we will continue to see large gaps between the buyer and the seller.
     

  3. TRAINING: Our AI researchers and interns will use these redlines to train an LLM to propose creative win-win solutions that go beyond what are in these redlines.
     

  4. FEEDBACK: We will then ask our contributors to rate (1) whether the AI proposed "creative win-win" solutions make sense, and (2) whether a redline created by LLM is better or worse than that created by an expert attorney (of course not knowing which one was created by AI).

Benefits of this project to our contributors

I hope this project provides an intriguing, fun and rewarding learning experience even for the most experienced attorneys. As you will see in the Google Drive, not only do our perceived "market" positions differ greatly, but our drafting does as well.

 

During the summer we will host two discussion/negotiation sessions via Zoom with your fellow contributors. It will be a great opportunity to meet your fellow experts from different companies and industries. Last but not least, if we can leverage AI to come up with better win-win solutions, that will be a benefit to our entire society. 

How else can I help?

Contact

Once you're done with your redline, submit it at the top of the page or at the following link: https://forms.gle/HzRe77WA1EhSiJo36. Thanks in advance for your contribution to the dataset!

If you're interested in working further with Atticus or RIDE, send us an email at info@atticusprojectai.org!

bottom of page