Confidential business information (CBI) shows up in many forms: nonpublic contracts, policy drafts, product roadmaps, pricing structures, or even a throwaway line in an email. It’s critical to catch, yet it’s rarely labeled, making it incredibly challenging to uncover among multi-page attachments or within lengthy communications.
Identifying this material often requires painstaking, line-by-line review across documents to catch any sign of sensitivity. And with 21st Century data volumes, the risks of getting it wrong keep rising.
Traditional workflows simply weren’t built for this nuance or amount of data. That’s why a new CBI analysis type in Relativity aiR for Review is now available, empowering you to use generative AI to surface confidential business content with greater speed, consistency, and confidence.
Why is CBI So Hard to Identify?
Compared to other review tasks, identifying CBI during review has distinct challenges that make it particularly complex:
- CBI is rarely labeled and takes many forms.
- Terminology varies across industries, organizations, agencies, business units, and even specific projects.
- Protective orders, confidentiality frameworks, or disclosure standards are often abstract, making it difficult for reviewers to understand what confidentiality level to assign to documents. Reviewers also often do so without indicating why, or what type of sensitive information a document contains.
- A single sentence can carry contextual meaning that’s only clear when compared against the larger data set. Information that might be fine to share in isolation can be considered confidential when grouped with other documents.
- The stakes are high: one missed document can expose business plans, while over-redaction slows review teams and can cause credibility issues with regulators, courts, or opposing counsel.
The takeaway? Traditional CBI review is difficult, time consuming, and expensive. Reviewers, with limited and varying levels of case knowledge, must make judgment calls that don’t scale, under tight timelines. Consistency feels nearly impossible, and when consistency slips, so does the protection organizations expect.
Introducing CBI Analysis in aiR for Review
CBI analysis in aiR for Review is uniquely positioned to solve CBI challenges. It combines the power of generative AI with structured reviewer guidance to deliver high quality, defensible classifications at scale.
This specialized approach offers distinct advantages.
- Fewer false alarms, fewer misses: aiR for Review’s CBI analysis flags what’s confidential based on content type, not gut instinct. It shows you the type of sensitive information found in the document and provides a direct citation for easy verification.
- Stronger compliance posture: Apply accurate, defensible confidentiality calls that stay aligned with protective orders, statutory obligations, and regulatory or policy expectations.
- Time saved, costs reduced: Cut down the hours spent on manual review and QC. Let aiR for Review handle first-pass CBI calls so you can establish more effective second-pass workflows, keep projects moving, and avoid expensive rework.
- Expertise where it matters: Leverage aiR for Review's decisions and reasoning to isolate the most confidential material so you can allocate subject matter or attorney expertise where it has the greatest impact.
Putting CBI Analysis into Practice
To use this new capability, review teams follow a clear, step-by-step process designed to maximize accuracy and efficiency.
1. Choose the Documents to Analyze
Reviewers begin by identifying the documents they want to review for confidentiality, whether that is a production set or a more specific selection of potentially sensitive documents. A sample from that population should be identified to help iterate on a prompt.
2. Select CBI Types Relevant to the Case
Reviewers choose the built-in CBI types—such as Trade Secrets, Technical Information, Financial Data, et cetera—to run on their documents. These selections signal to aiR what needs to be flagged and how to focus on the right patterns.
3. Add Case-Specific Context
Every case, investigation, or regulatory request involves unique circumstances. Reviewers can add case-specific context to further refine each CBI type and help aiR for Review better understand what should be considered confidential information for the specific matter at hand.
4. Map CBI Types to Existing Confidentiality Coding
aiR for Review gives users the control to select confidentiality designations based on their review needs and processes. They map field choices like Confidential, Highly Confidential, Counsel's Eyes Only, or whatever else they use as a confidentiality designation in their workspace to the different CBI types being assessed. aiR for Review then uses this context to appropriately designate documents based on the CBI they contain.
5. Run the Analysis
Once setup is complete, users conduct CBI analysis on their sample documents, evaluate initial results, iterate on prompt criteria as needed, and, finally, run the analysis at scale when ready.
CBI Analysis, Put to the Test
Cimplifi, an early adopter of Relativity aiR, was a member of the Advanced Access program for CBI analysis. After experiencing how aiR for Review accelerated responsiveness and issues reviews, they were eager to explore how a dedicated CBI workflow could elevate their work even further.
“We’ve seen the impact of aiR for Review’s other analyses, so when we had the chance to try out CBI analysis early, it was an easy yes.”
- Shannon Mazur, Senior Review Consultant, Cimplifi
Cimplifi ran aiR for Review’s CBI analysis on over 40,000 documents that required rapid corrections related to confidentiality designations. Working under a tight timeline, Cimplifi ran the documents through aiR for Review to help reassess those calls and finalize productions. The results gave their team immediate clarity, highlighting specific CBI types in each document and indicating why the model reached each prediction. This helped them establish consistent, defensible confidentiality decisions that satisfied both client and court expectations in far less time than an alternative, manual approach.
“We needed clarity on tens of thousands of documents that had been too broadly assessed for confidentiality. With a strict deadline pending, the overinclusive designations were causing issues. We turned to aiR for Review to better identify the confidential material with supporting rationales and citations. It allowed us to immediately downgrade half of the population and focus reviewer time on more complicated calls. We were able to reproduce in very short order and the client was thrilled.”
- Shannon Mazur, Senior Review Consultant, Cimplifi
Greater Impact on Review; Better Protection for Sensitive Data
Confidential business information is essential to find and protect, as it can reflect an organization’s strategy, innovations, pricing philosophy, or competitive posture. That’s the type of information you don’t want to accidentally send out the door. Yet until now, identifying it has been one of the most time-consuming and error-prone parts of review.
With CBI analysis in aiR for Review, legal teams can finally approach this challenge with modern tools—reducing risk, increasing consistency, and accelerating review without compromising accuracy. It’s a smarter, more defensible way to protect what matters most.
Interested in learning more? Reach out here to request an aiR for Review product demo.
Graphics for this article were created by Caroline Patterson.
