December 8, 2025 - 5 min read

Keep Classification Aligned as Your Organization Grows

Arctera
Headshot of Chris Stapenhurst, Director, Product Management

Chris Stapenhurst

Director, Product Management

I’ve met with many of our customers worldwide this year, and the conversations all point in the same direction: classification programs are expanding rapidly but maintaining them hasn’t gotten any easier. When departments reorganize or new communication channels surface, policies don’t always keep pace. And when auditors look for justification, teams need classification logic that is transparent, explainable, and defensible.

That’s exactly what guided this release of Arctera Insight Classification. We set out to automate the manual work that burdens teams today, expand our transparent policy framework, and deliver tools that help you fine-tune accuracy and reduce false positives. Ultimately, my goal is to help your teams stay ahead of organizational and regulatory change, while improving confidence in every classification outcome.

Auto-Update for Department-Based Policies

One of the most common requests I hear from teams is for classification to stay aligned with their actual organizational structure without requiring constant updates. Many of you rely on the Author Department and Receiver Department conditions to maintain precision, but when departments shift, policies quickly fall out of sync.

Screenshot of Arctera Insight’s Department selection interface showing automated department updates for classification policies, including child and root department controls.

With this release, we’ve solved that problem. Arctera Insight Surveillance department hierarchies now sync automatically to Classification every day. New or removed departments are reflected in your policies without manual effort.

When Auto-Update is enabled on a parent department, all child departments come along with it. And if it's turned on at the root level, any new departments created there are added automatically. This ensures your policies stay aligned with real-world organizational change and reduces the operational load on your teams, especially during periods of growth or restructuring.

Example of Auto-Update activated in Arctera Insight, displaying selected departments and automated updates to maintain classification coverage as organizational structures change.

Strengthening Transparency with Enhanced Policies

Another theme I consistently hear from customers is the growing importance of transparency. Whether for internal review, external audits, or regulatory inquiries, teams need complete clarity into why a message was classified a certain way.

Our transparent policies were built for precisely this purpose, and in this release, we’ve expanded them even further. Every rule and pattern remains fully visible and editable in the console, giving you greater defensibility and control.

These updates were shaped directly through client collaboration:

  • Client Concerns: Now includes terminology associated with “ghosting.”
  • Secrecy: Expanded with additional terms to strengthen detection.
  • Off-Channel Signaling: Enhanced with new terms, abbreviations, and communication channels to improve coverage.

These updates reinforce our commitment to transparent, explainable classification logic. They help reduce noise, improve accuracy, and give teams greater confidence in how policies behave across real-world communication scenarios.

Additional Enhancements in This Release

Enhanced Analyze & CSV Export

A common challenge we hear from teams is the need for deeper visibility into what triggered a classification result, especially during testing and model tuning. To support that, we've introduced a more detailed CSV export that adds fields such as Highlighted Text, Matched Text, and Policy Name.

These additions make it easier to review triggering sentences, understand pattern behavior, and validate policies before you promote them into production.

Screenshot of Arctera Insight’s enhanced CSV export showing highlighted text, matched text, and policy names used to analyze classification accuracy and tune detection patterns.

New Policies

To help customers stay ahead of evolving regulatory requirements, we’ve expanded our policy library with several new additions:

1) Disclaimers Policy – Designed to identify new or modified disclaimers that often lead to false positives. These can be incorporated into workflows as index exclusions for Surveillance or eDiscovery.

2) Jurisdiction-Specific Privacy Policies:

  • Minnesota Consumer Data Privacy Act (CDPA / “Minnesota Act”)
  • Maryland Online Data Privacy Act (MODPA) and Maryland Personal Information Protection Act (PIPA)
  • Nebraska Data Privacy Act (NDPA / LB 1074)

These policies strengthen coverage for fast-changing privacy expectations and help teams maintain more reliable, compliant classification programs.

Looking Ahead

As classification programs grow in scope and complexity, our focus remains the same: deliver automation, transparency, and accuracy that meaningfully reduce operational burden while improving trust in AI-driven detection. This release moves us further toward that goal.

I appreciate the feedback and partnership from so many of you who helped shape these enhancements. As always, we’ll continue to build with you—and for you—as classification becomes even more central to managing regulatory, privacy, and behavioral risk.