Rhindon Cyber logo
    Support & Documentation
    Contact Support

    Release Notes

    What shipped in RAIC, newest first. Dates reflect production availability on app.rhindoncyber.com. Feature availability varies by tier — see the tier notes on each module page.

    Admin downloads, draft deletion and webhook signing

    Most recent release (November 2026).

    • Fixed: the CIS Controls and NIST CSF assessments, the CIS Controls page and the CIS Controls report could show "Catalog is empty" with no controls to score. The controls are back, and answers already saved, including those filled in by the Cross-Framework Questionnaire, show again.
    • New Assessments page under Compliance lists every framework's assessments with its latest one. An AI / Cyber switch on that page and on the Cross-Framework Questionnaire, like the one on the Compliance Scorecard, shows only the cyber frameworks (NIST CSF 2.0, CIS Controls v8.1, SOC 2, HIPAA) and the company-wide cybersecurity modules, or only the AI side.
    • Cross-Framework Questionnaire: new company-wide cybersecurity modules ask about your whole organization (asset and vulnerability management, identity and data protection, training, logging, incident response, backups, suppliers, advanced technical controls, governance, physical security, privacy and processing integrity). Together they cover every item of NIST CSF 2.0, CIS Controls v8.1 (IG1 to IG3), SOC 2 (including the optional criteria) and, with HIPAA mode on, the HIPAA Security Rule, and each answer fills every framework it maps to at once. The seven AI modules are shown in their own section.
    • Supplier Document Intelligence can now read supplier PDFs with AI. Turn on "Read supplier documents with AI" in a supplier's Files panel (off by default): each PDF is read for its report type, dates, audit opinion, certifications, subprocessors and retention terms, and the agent flags out-of-date reports, audit exceptions, documents about another company, new subprocessors and changed retention terms. Files are sent to Google's Gemini API, which does not train on them and keeps requests for up to 55 days for abuse monitoring.
    • POA&M: each roadmap's Export menu now downloads OSCAL 1.1.2 JSON for auditors, regulators and government customers. Choose POA&M for the roadmap as an OSCAL Plan of Action and Milestones, or Assessment Results for your status on every control of the roadmap's frameworks.
    • Regulatory Intelligence: each law in the Regulatory Register shows its stage (Proposed, Enacted, not yet effective, or In force) and a Brief with an overview, how it applies, the risks it brings by stage and a clock to its enforcement date. Proposed bills no longer raise suggested risks or deadline emails, muted laws raise none, a law now applies only where its jurisdiction reaches you, and saving your regulatory profile re-evaluates what applies.
    • Regulatory Intelligence: press releases and announcements now appear as News, with no requirements and no suggested risks. Each law's brief now explains in plain words how it applies to use cases, AI systems and AI controls, and lists the steps left for proposals that are not state bills.
    • New Admin screen, Where We Operate: pick the regions and countries you operate in (the European Union is one choice) and, for the United States, all 50 states in one click or the individual states. State laws then apply only where you operate. It is also a new step on the Getting Started checklist.
    • Regulatory Intelligence: Where We Operate lists US cities with their own AI laws, such as New York City. When a law for a city not covered before is published, Org Admins get a bell notification, an email and a pop-up asking them to check whether they operate there.
    • Regulatory Intelligence: every morning RAIC re-evaluates what applies to you, so new or changed AI systems and use cases are picked up within a day without pressing Re-evaluate. A new version of a law takes over on its effective date, proposed bills that are vetoed or fail drop out of your register automatically, and repealed or superseded laws stop raising suggestions. An AI system's annex pack now lists the regulatory risks you accepted for it.
    • Regulatory Intelligence now watches the US Congress: House and Senate bills about artificial intelligence are collected every day from Congress.gov and reviewed before they reach you, and a federal bill's brief lists the steps left in the House, the Senate and with the President.
    • Org Admins can download the RAIC app package for Teams and Copilot, and the latest RAIC Teams App Install Guide, under Admin → Integrations, instead of asking Rhindon support for them.
    • Admin → Integrations → Microsoft AI Discovery now starts with the RAIC agent in Microsoft Entra panel, since the agent can automate the integration. Just below it, Org Admins can open the latest RAIC Agent Registration Guide, with step-by-step instructions for connecting the RAIC Governance Agent to your Microsoft tenant.
    • New Support guides: Regulatory Intelligence, NIST AI RMF 1.0 (with the NIST Agentic Profile), CIS Controls v8.1 and Email Delivery.
    • Partner Portal: Managed Tenants shows 25 tenants a page with a pager, and lists every tenant even past 1,000.
    • Header: every member now sees their organization's name, NFR badge and trial end date in the top bar; before, only Org Admins did.
    • Use Case Register: OrgAdmins, WorkflowAdmins and Super Admins can delete a draft use case (Pending Triage, not yet submitted or decided) from Lifecycle Actions on its detail page. Use cases further along are retired instead.
    • AI System Registry and Controls: the same admins can delete a draft AI system (Pending Triage) from Lifecycle Actions, and a draft control from its page or the controls list. Submitted or approved records can no longer be deleted.
    • API webhooks: endpoints can now subscribe to governance policy events (policy.created, policy.updated, policy.deleted), and the API Reference now gives the exact signature check: the HMAC key is the SHA-256 digest of your whsec_ signing secret.
    • API webhooks: new endpoints are signed with your whsec_ signing secret itself, which RAIC now keeps encrypted, so the stored digest of the secret can no longer be used to forge a signature. Endpoints created before this update keep the digest key until you delete and recreate them; the API Reference shows both checks.

    Fixes

    Fixes in recent releases.

    • New AI System Registration: applying an AI catalog template now fills System Type and Deployment Model with the matching value from your organization's lists (for example Cloud (Public) picks Cloud - Public) instead of leaving the box blank. When your organization has no system types, deployment models or business processes defined, the page now says so under the dropdown, with a link to Admin → Tables for Org Admins.
    • Organizations that were missing their System Register lists or alert settings have been given the defaults. Lists you had already filled in were left as they were.
    • New organizations created or cloned by Rhindon Cyber, and Partner demo tenants, now start with the default System Register lists (system types, deployment models, business processes, regulatory frameworks and vendor certifications) and alert settings, like every other new organization.
    • Cross-Framework Questionnaire: typing an evidence note is no longer slow and no longer drops letters. The note saves a moment after you stop typing and at once when you leave the field, and Submit waits for any answer still saving.
    • Assessment wizards: Rhindon Cyber support staff working in your organization can now create and update SOC 2, CIS Controls, HIPAA, NIST CSF and NIST AI RMF assessments; creating one stopped with a row-level security error. Capture snapshot in the NIST CSF wizard also works again.
    • POA&M CSV exports now name each framework control a task covers (for example soc_2 CC2.1) instead of showing [object Object] for generated roadmaps.
    • Connecting Ask RAIC or another app to RAIC no longer stops on "Auth session missing!" when the sign-in window holds an expired RAIC session: the consent screen clears it and asks you to sign in again.
    • Partner extension rollout no longer offers a Safari download: Safari installs extensions only from the App Store, so the Safari tile reads App Store listing coming soon until the listing is live.
    • New Partner firm admins and invited Partner staff can sign in: they no longer get "No account profile found", and setting their password from the setup email activates their firm membership. Rhindon Cyber now sets a new Partner firm's license (Enterprise by default) when creating it.
    • Training Register: the course assignment pickers now list every active user in your organization; users who were not OrgAdmins used to see only themselves.
    • NIST CSF POAMs: if the Subcategory response for an evidence upload cannot be loaded, the register now says so instead of "No Subcategory response yet".
    • Marketplace: if the listings cannot be loaded, the page now says so instead of "No marketplace listings are published yet".
    • A page opened in a tab from before an RAIC update now reloads once to the latest version instead of showing "This section didn't load".
    • Approval delegations: a delegation whose end date is before its start date is now refused with a message instead of being saved.
    • Settings: if your email notification preference cannot be loaded, the page now says so instead of showing it as off.
    • FINRA WSP register: if a review cannot be saved, the page now says so instead of showing "Review recorded" and updating the last review date.
    • Docs buttons: the Docs buttons on the Formal Supplier Register, GPAI Provider workflow, Trust Portal, internal intake forms, AI Spend Report and Recertification Pack open their documentation instead of a page-not-found screen, and the EU AI Act and ISO 42001 Annex A assessment wizards open their guides.
    • MCP Server: the Docs button now opens the MCP Server support page, and the MCP tools for AI Catalog submissions and AI Data Governance changes now work for callers holding those permissions.
    • Sign-in: right after you enter your authenticator code, the home page, the Terms of Service check and guided tours no longer show errors while two-step sign-in finishes.
    • Messages: form checks, plan limits and refusals (for example a missing required field, or a feature your role cannot use) now show as warnings instead of errors, with the same wording. Every failure shown as an error is now also recorded for Rhindon Cyber support.

    Governing Microsoft agents

    October and early November 2026.

    • Regulatory Intelligence: new regulations are now cataloged with the specific obligations they create, drafted with AI and checked by Rhindon Cyber staff before they reach your register.
    • Regulatory Register: when a regulation applies to you and nothing covers it, the register suggests a risk and a remediation task for an OrgAdmin to accept or dismiss.
    • Microsoft AI agents (Copilot Studio, Foundry, Agent Builder and Agent 365) found by Microsoft 365 discovery are labeled in the NHI registry, and the ones not yet in your register go to the Shadow AI Triage agent, which can draft an AI System intake.
    • Agent drift compares each approved agent with what Microsoft reports: agents running but not approved, approved agents blocked or deleted, sponsors who are not owners, and new permissions.
    • Optional Agent 365 registry connector lists every registered agent, including declarative Copilot agents with no Entra identity.
    • Microsoft Agent Enforcement proposes blocking an agent in Microsoft when its RAIC approval lapses, and unblocking it once approved again. Proposals only: you make the change in the Microsoft 365 admin center.
    • Auto-Risk Drafts score each Microsoft agent on autonomy, data reach and EU AI Act tier, and promoting a draft creates a draft impact assessment.
    • Microsoft evidence: daily, dated readings of seven Microsoft 365 settings (MFA, legacy authentication, Conditional Access for agents, sensitivity labels, Secure Score) mapped to ISO 42001, NIST AI RMF and CIS.
    • Optional Microsoft agent alerts and activity: Defender alerts that name an AI agent open incidents, and Agent 365 activity feeds the NIST Agentic Profile.
    • ISO/IEC 42001 Annex A is mapped control by control to Microsoft Purview Compliance Manager, with a CSV download.
    • Microsoft 365 Snapshot: a free, read-only scan of your tenant scored against ISO 42001 and the NIST AI RMF, with a Word gap report. Partners run it from a Shadow AI Discovery engagement.
    • RAIC in Teams and Copilot: submit AI use cases and decide intakes on cards in Teams, get approval alerts there, and ask questions with the Ask RAIC agent in Microsoft 365 Copilot.
    • Systems RAIC discovers no longer count toward your plan's AI system limit until they are approved.
    • Board One-Pager PDF: the AI and Cyber maturity Board One-Pager downloads as a one-page PDF with the headline score, change since the last snapshot, each dimension against its 6-month target, top initiatives and framework coverage.
    • Partner QBR PDF: partners can export the Quarterly Business Review as a designed PDF, alongside the editable Word version, using their white-label name.
    • Branded PDF exports are much smaller: the logo is now compressed, so a typical report is a few hundred KB instead of about 5 MB.
    • Ask RAIC can run your startup checklist: an OrgAdmin asks it to finish setup and it goes through the Getting Started steps one at a time (allowed email domains, email deliverability, System Register lookups, governance designations, use case register settings, alert and review cadence). Inviting team members stays in RAIC.

    Now hosted on Microsoft Azure

    • RAIC moved to Microsoft Azure (East US 2) on October 23, 2026, behind Cloudflare. Your data stays in the United States.
    • Company sign-in now skips RAIC's own authenticator code only when your identity provider shows it completed multi-factor sign-in; Microsoft Entra ID needs the amr claim (see the SSO guide).
    • Passwords must be at least 14 characters with upper and lower case, a number and a symbol, and passwords found in known breaches are refused.
    • MCP connections from ChatGPT, Claude, Cursor and Microsoft 365 Copilot work again, with permissions chosen on RAIC's consent screen and read-only by default.
    • Shadow AI browser extension v2.3.0 runs on Firefox 140 and later. Firefox asks you to agree to the data the extension collects (browsing activity, website activity and website content) when you install it. Chrome, Edge and Safari behave the same as v2.2.2.
    • Firefox users and MSPs now download a Mozilla-signed Firefox add-on (.xpi) from Settings and the MSP portal. It installs permanently, updates itself, and MSPs can force-install it with the Firefox policies.json shown on the Extension Rollout page.
    • Generating an organization extension key in Org Settings now also shows a Firefox policies.json with the key filled in, so organizations without an MSP can force-install the Firefox add-on through Group Policy or Intune.
    • Shadow AI browser extension v2.2.2 no longer logs a false tamper warning (HMAC envelope mismatch) when it switches to a different person's account while a settings download is still in progress.
    • Shadow AI browser extension v2.2.1 opens the Connect Browser Extension page in a new tab when a different person signs in to RAIC, so the switch also works for an account that has never connected the extension before.
    • Shadow AI browser extension v2.2.0 switches to whoever signs in to RAIC in the same browser, uploading the previous account's detections first; devices set up with an organization key or by your IT team never switch. The Check this page debugging button is removed from the popup.
    • Shadow AI browser extension v2.1.2 shows the unapproved-AI warning on the first visit in Safari on iPhone, not only after a reload, and uploads detections when a visit is recorded and the last upload is over 30 minutes old, because iPhone Safari doesn't run the extension's sync timer.
    • Shadow AI browser extension v2.1.1 stops Safari from showing the "Allow on all websites" notice when all websites are already allowed, and adds Check this page to the popup, which shows why the unapproved-AI warning does or doesn't appear on the open page.
    • Shadow AI browser extension v2.1.0 shows the unapproved-AI warning reliably in Safari on iPhone, and its popup now says when the browser hasn't allowed it on every website, with the steps to allow it. It asks for no new permissions.
    • Shadow AI browser extension v2.0.1 adds the large app icons and the name Shadow AI Discovery needed for the Safari App Store listing. Chrome, Edge and Firefox installs behave the same as v2.0.0.
    • Shadow AI browser extension v2.0.0 trusts only rhindoncyber.com addresses now that the old hosting addresses are retired. Download it again to update an existing install; it asks for no new permissions.

    Governance Operations and Decision Rights

    • Governance Operations now has its own workspace and menu, separate from Compliance, opening on the Operator Cockpit.
    • Operator Cockpit: one prioritized queue of everything due or overdue, with a plain-language explanation of why each item is ranked where it is.
    • Councils, meetings, agendas and minutes, with decisions written to a Decision Register linked back to the forum and the records concerned.
    • Action Register, exceptions and waivers with expiry and renewal, an escalation ladder, and a governance calendar you can export to your own calendar.
    • Authority Matrix: define who may approve what, propose and publish versioned rule changes, and see how many existing records a change would reroute before it takes effect.
    • Approval steps on use cases, AI systems, controls and policy versions are generated from your matrix and fail closed, refusing self-approval and approvals beyond a delegation limit.
    • Risk acceptance as its own decision right, with a register, expiry and reassessment, plus time-bounded emergency authority that must be closed out.
    • Governance State and Continuity Pack: a branded briefing, every register, a calendar file and content hashes for 90 days, 12 months or all time.

    Evidence Integrity and the Trust Portal

    • Verifiable evidence lineage: governance history is hash-chained and sealed nightly, so a record cannot be changed after the fact without the chain showing it.
    • Signed audit packs an external auditor verifies without a RAIC account, using a bundled offline verifier and a published specification.
    • Trust Portal: a public, link-gated governance page for customers, prospects, and auditors, with per-section visibility, access requests, and visit logging.
    • Pre-publish evidence check on the Trust Portal — publishing is blocked only when the evidence chain itself would fail verification, and advisory findings are recorded with your acknowledgement.
    • Portable continuity badge you can embed on your own website, backed by a discoverable descriptor and a redacted public signal.
    • Independent verification attestation published on a fixed quarterly cadence.

    Agentless vendor collectors

    • Read-only, detect-only connections to Microsoft 365 Copilot, ChatGPT Enterprise, and Claude (Enterprise/API) — no endpoint agent, and no prompt or file content leaves the vendor.
    • Claude is collected as two independent views, administrative activity and usage analytics, each on its own schedule so a slowdown on one never blinds the other.
    • Unusual jumps in usage volume are flagged for review against your own recent baseline.
    • Vendor credentials are stored encrypted, and every sweep is recorded in your activity log and the platform log.

    AI Data Governance

    • Inline inspection of content flowing into AI tools, with classification by pattern matching and named entity recognition.
    • Allow, mask, or block policy applied inline or retrospectively, per data class.
    • Encrypted payload vault so sensitive captures are retained for investigation without being readable in the clear.
    • Findings feed the Shadow AI workspace and the risk register.

    Automation Agents

    • Governance agents for Shadow AI triage, evidence freshness, and related routine work, under Governance → Automation Agents.
    • Four autonomy modes per agent — off, draft only, auto-act, and fully autonomous — with a confidence threshold you set.
    • Run an agent on demand or leave it on its schedule; every action is attributed and audited.
    • New in-app how-to video walking through enabling, running, and reviewing agent results.

    Anonymous peer benchmarks

    • Compare governance metrics against anonymized peers in your industry sector, size band, and region.
    • Cohorts are published only once a sufficient number of firms contribute, so no single organization can be identified.
    • Cohort tags are assigned by Rhindon Cyber rather than self-declared, keeping comparisons trustworthy.
    • Peer position appears on the maturity dashboard, registers, and executive reports.

    Partner template library

    • Install canonical RAIC document templates straight into your firm's Partner Docs library from the MSP portal.
    • New MSP Shadow AI Discovery Package template for client engagements.
    • Re-installing an unchanged template is a no-op; updates land as a new document version.

    Platform videos and documentation

    • Videos section in the documentation hub: stream the MSP Partner Portal demo, MSP onboarding, the General, Starter, and Professional client onboarding walkthroughs, and the Automation Agents how-to in-app.
    • Support site rebuilt on a canonical documentation registry, so every page now appears in the index, the sidebar, the sitemap, and the AI-crawler files from one source.
    • Twelve new documentation pages covering the Risk Register, EU AI Act workbench, NIST CSF 2.0, HIPAA, Hedge Fund / FINRA WSP, AI Discovery, AI Maturity, MSP Partner Portal, MCP Server, API Tokens, Core Concepts, and these Release Notes.
    • OWASP Top 10 review completed: security response headers published, leaked-password checks enabled at sign-up, and SECURITY DEFINER function grants swept.
    • Dependency vulnerability scanning runs on every build.

    AI Catalog and Curator pipeline

    • Curated AI catalog expanded past 1,500 entries across web, model, and API detections.
    • Automatic promotion of high-confidence unknown signals into catalog drafts, with automatic patch application for safe fields.
    • Multi-tenant host handling so shared platforms no longer collapse distinct tenants into one row.
    • Version history now supports a side-by-side comparison against the previously published version.
    • Pending-change viewer showing field-level differences before you publish.

    Risk and maturity

    • Dual inherent and residual risk scoring across AI Systems, Use Cases, and the Risk Register, with a heatmap toggle between the two views.
    • AI Maturity Index dashboard with GMI, CGMI, and TGP scoring plus daily snapshots and trend charts.
    • PowerGRYD pyramid reporting: overview, maturity assessment, layer coverage, gap and risk heatmap, and trend analysis.

    Frameworks and compliance

    • HIPAA module, now in the Compliance and Suite packages or Core with HIPAA as its framework.
    • Hedge Fund / FINRA Written Supervisory Procedures overlay for financial-services firms.
    • SOC 2 Type 2 evidence module with the full AICPA Trust Services catalog and an auditor evidence pack.
    • NIST CSF 2.0 and CIS Controls v8 coverage through Cybersecurity Mode.
    • External ISO 42001 auditor workspace with engagement scoping, findings, and a branded report.

    Integrations and platform surface

    • RAIC MCP Server, letting AI assistants query governance data over Model Context Protocol with OAuth 2.1 and full audit logging.
    • Public REST API with org-scoped and platform-scoped tokens, webhooks, and rate limiting.
    • ConnectWise PSA connector for outbound risk-event push.
    • Microsoft Graph discovery for AI usage, entitlements, non-human identities, and Purview labels.
    • Shadow AI browser extension v1.18.3 for Chrome, Edge, and Firefox, with Intune and ADMX force-install support and MSP white-label branding for partner-managed clients.

    MSP and partner

    • MSP Partner Portal with a multi-tenant client launcher, demo sandboxes, and QBR reporting.
    • White-Label add-on for partner-branded portals, emails, and exports.
    • Guided onboarding tour for partners new to the platform.