September 2026
New Features
Section titled “New Features”Command Palette and Document Search
Section titled “Command Palette and Document Search”Press Cmd/Ctrl+K anywhere in the platform to open a new command palette. Type a risk or incident ID (for example RSK-42) to jump straight to it, or type free text to search across risks, incidents, and compliance documents at once — documents are now included in search for the first time. The palette also surfaces actions relevant to where you are, such as switching your theme.
Pressing / focuses the register’s search box directly, the same shortcut used by GitHub and Jira.
AI Assistants Can Read and Propose Changes to Compliance Documents
Section titled “AI Assistants Can Read and Propose Changes to Compliance Documents”AI assistants connected through the MCP server can now work with your compliance documents, not just risks and incidents. An assistant can list available document versions, read approved policy sections, propose edits — including organization-specific supplements — and review or approve/deny a pending proposal, using the same approval workflow as edits made in the document viewer. Only approved, published content is visible by default; a proposal’s contents stay private until it’s explicitly reviewed.
Post-Incident Review: One Setting for the Whole Decision
Section titled “Post-Incident Review: One Setting for the Whole Decision”The Cyber Incident Response Plan’s (CIRP) Post-Incident Review setting has been consolidated from two separate controls into one. Choose a single minimum severity — from SEV-5 and above down to SEV-1 only — or exclude the Post-Incident Review requirement entirely. Previously, two different parts of the document could each state a different rule for when a review was required.
Redesigned Policy and Procedure Document Exports
Section titled “Redesigned Policy and Procedure Document Exports”The Word document you download from a policy or procedure now opens with a proper title page — your organization’s logo and name, the document title, and its last modified/approved dates — followed by a table of contents that jumps to any section, and a body with numbered sections and lettered sub-lists. The footer is cleaner too: the Adversarial mark and “Powered by adversarial.com” on the left, the document name and page number on the right.
Improvements
Section titled “Improvements”One Work Item Editor
Section titled “One Work Item Editor”The centered work item editor, previously available as an opt-in Early Adopter preview from Settings > Preferences, is now the only editor for risks and incidents. The classic side-panel editor and the preview toggle have both been retired.
Governance Deck Tables Grouped by Urgency and Severity
Section titled “Governance Deck Tables Grouped by Urgency and Severity”Risk and incident tables in generated board decks and CyberGov reports are now grouped by urgency or severity band, with Closed items listed first within each band, and each table page captioned with the band it covers. A band with no matching records no longer produces an empty table page.
Faster Dashboard Loading
Section titled “Faster Dashboard Loading”The Home dashboard and the Remediation Agility (RemAgi) chart now load faster, particularly for organizations with large risk and incident registers.
Incident response-time figures — Mean Time to Respond on the Home dashboard and the incident chart’s average response time — may shift very slightly as a result: calculations now count fractional seconds instead of dropping them, and no longer treat a response clocked before an incident’s detection time as an instant response.
Calendar Document Version Numbers Are Now Fully Padded
Section titled “Calendar Document Version Numbers Are Now Fully Padded”Compliance document version numbers now always display with two-digit month and revision numbers (for example, 2026.09.01), rather than mixing single- and double-digit forms.
Broader MCP Protocol Support
Section titled “Broader MCP Protocol Support”The MCP server now also supports the 2026-07-28 revision of the Model Context Protocol, alongside the version it already served, so AI assistants built against the newer specification can connect without issues.
Tagging Risks and Incidents at Creation
Section titled “Tagging Risks and Incidents at Creation”The API used to create risks and incidents now accepts a list of existing tags to apply at creation time, for integrations and automations that create fully-formed records in one step.
- Pasting a multi-line cell from Excel or Google Sheets into a compliance document editor is now supported.
- If your session expires, you’re now taken straight to the sign-in page as soon as the platform detects your session has ended.
- The organization picker on the integration consent screen now scrolls to account for multi-organization users, rather than push content off screen.