Skip to content
fullCircle GRC

Product update

New in fullCircle: June 2026

AI agents for vendor reviews

June put AI agents to work on vendor risk, taught the Evidence Collector's Autopilot to take direction mid-run and made it easy to build one assessment across several frameworks. It also opened early access to connecting fullCircle to Claude.

Share

This month you can:

  1. 01AI agents that run your vendor reviews
  2. 02Steer Autopilot without starting over
  3. 03One assessment across several frameworks
  4. 04Connect Claude to fullCircle, in early access
+ 8 more improvements
VendorsEvidenceAuditsMCP Server
01Vendors

AI agents that run your vendor reviews

Vendor reviews pile up: scheduling them, sending questionnaires, reading SOC 2 reports. Two new AI agents take that on. Generate Reviews creates upcoming reviews and sends the right questionnaires. Review Vendors reads vendor reports, summarizes findings and flags potential risks. You get an alert when a run finishes.

Generate Reviews from the vendor list or side card
Review Vendors summarizes reports and flags potential risks
An activity log for every run
Email or Slack alerts when a run finishes or fails
Explore Vendor Risk Management

Part of vLM in xLM Agent Suites, added to Optimize and quoted per suite.

02Evidence

Steer Autopilot without starting over

Automation is only useful when you can correct it mid-flight. The Evidence Collector's Autopilot now takes guidance while it runs: edit its plan, skip a task or add a hint. It sends borderline screenshots to you for a decision, types into search and filter boxes and captures true full-page screenshots.

Add guidance and edit the plan while Autopilot runs
Borderline captures sent to you to decide
Types into search and filter boxes to find evidence
Plans from evidence you've accepted before
Explore Evidence Automation

Included with Optimize.

03Audits

One assessment across several frameworks

Testing the same control once per framework wastes everyone's time. Generate Assessment now combines several frameworks: controls they share become a single testing entry that keeps every framework reference, and tags from your controls and evidence come along. New tag filters include or exclude tags with AND or OR logic.

Generate one assessment from multiple frameworks
Shared controls tested once, with every framework reference kept
Tags carried over from controls and evidence
Tag filters that include or exclude, with AND or OR
Explore Audit Management
04MCP Server

Connect Claude to fullCircle, in early access

Your team already asks AI assistants for help. Now those assistants can see your program. Early-access customers can connect Claude to fullCircle's MCP Server with OAuth, sign in with their existing fullCircle login and switch between organizations right from the chat.

OAuth sign-in with your fullCircle login, including SSO
Switch organizations from Claude
New tools for assessment controls and customizations
Explore the MCP Server

Included with Optimize.

Also in this release

Smaller changes that make everyday work smoother.

QuestionnairesQuestionnaire Responder adds questionnaire names, a confidence threshold for auto-filled answers and better question detection
AuditsJump from an audit request to its linked control evidence
AuditsFind Existing Evidence links files to controls that have no mapped request
PlatformLinks to another organization's records prompt you to switch organizations
ControlsReorder saved views on Controls
RiskRisk register bulk uploads accept scopes
PlatformA refreshed dark mode across charts, AI chat, the calendar and the policy editor
AuditsScoping Factors for internal assessments: yes-or-no questions that scope controls in or out, with a default narrative

See June's release in action.

Book a demo and we'll show you what's new, using the frameworks and workflows your team runs today.

Or see how audit-ready you are