Release Notes - AIR

Binalyze AIR 5.30

Written by Elif Kurt | Tue, Nov 10, '26

What’s New?

  • Introducing Automations: AIR now enables teams to automate investigation and response workflows based on events, schedules, or requests from external systems. With a visual workflow builder, users can automatically initiate threat hunts, collect evidence, isolate assets, apply auto tagging rules, and send contextual notifications to email, Slack, or external security tools. Reusable predefined templates, flexible asset targeting, and detailed execution history help teams reduce manual effort, accelerate investigations, and ensure consistent responses across their environments. For configuration details and supported capabilities, visit the Automations Knowledge Base.

  • LDAP and Active Directory role mapping: AIR can now map directory groups to AIR roles during login, provision LDAP users automatically, refresh identity fields from the directory, and clean up removed or disabled directory users. This helps administrators keep access aligned with enterprise identity governance.

  • Resource Limit profiles: Task resource controls are now organized into predefined levels such as Minimum, Low, Balanced, High, Maximum, and Custom. Analysts can quickly choose an appropriate resource profile while still seeing and editing the underlying limits when needed.

  • Improved update controls for SaaS: Global Admins can now view pending update requests, defer updates, schedule updates for a specific time, request an immediate update, and see update status in the redesigned AIR Update Settings experience.

  • Expanded browser artifact coverage: Acquisition profiles now include additional browser artifact sources such as Firefox and Safari bookmarks, Firefox and Safari favicons, and Chromium omnibox shortcuts. These sources provide additional context about user intent, visited pages, and browser activity during investigations.

New Features & Improvements

Automations

Introducing Automations: Streamline Investigation and Response Workflows

AIR now introduces Automations, enabling security teams to build automated investigation and response workflows that reduce manual effort, accelerate operations, and ensure consistent execution. Using an intuitive visual workflow builder, users can configure automations triggered by events such as asset registration, completed acquisitions or threat hunts, isolation status changes, threat intelligence updates, scheduled times, or HTTP requests from external systems. Flexible conditions and asset targeting allow workflows to run only when specified criteria are met.

Each automation can execute multiple actions in sequence, including starting threat hunts, collecting evidence, capturing baselines, applying asset tags, isolating or unisolating devices, and sending notifications via email or Slack. Automations can also communicate with external security tools through HTTP requests or send contextual data directly to SIEM platforms. Notifications and outbound requests can be customized using dynamic variables from triggers, previous actions, and their outputs, enabling teams to share relevant investigation information automatically.

To simplify workflow creation, Automations provides predefined templates, alongside the ability to create custom workflows, duplicate existing automations, and import or export configurations for reuse. Execution history, configuration change tracking, and role-based access provide visibility and control over automated operations. Together, these capabilities help security teams minimize repetitive tasks, respond faster to security events, and integrate AIR more effectively into their existing security operations workflows.

For configuration details and supported capabilities, visit the Automations Knowledge Base.

Organization Administration

Redesigned Update Settings for SaaS

Global Admins can view pending update requests, defer or cancel updates, schedule updates for a specific time, and request an immediate update. Update status is shown on the redesigned Update Settings page.

LDAP and Active Directory Role Mapping

Administrators can now map Active Directory / LDAP groups directly to AIR Console roles, enabling centralized, directory-driven access control. When users log in with their domain credentials, AIR automatically evaluates their active directory group memberships and assigns the appropriate role dynamically, without requiring local user accounts to be pre-created or maintained in AIR. Any joiner, mover, or leaver updates made in corporate directory (such as role adjustments, group changes, or account deactivations) immediately take effect upon login, eliminating manual user administration in AIR and preventing access drift.

Evidence Collection and Asset Workflows

Expanded Browser Artifact Collection

Predefined acquisition profiles now include additional browser artifact sources. Firefox bookmarks, Safari bookmarks, Firefox favicons, Safari favicons, and Chromium omnibox shortcuts are now available in supported profiles.

These sources help analysts understand saved destinations, accepted address-bar completions, page-to-icon relationships, and corroborating browser activity. The added sources improve evidence coverage for user activity investigations across Windows, Linux, and macOS.

Evidence Staging Notes for Repository Uploads

When analysts configure evidence collection to save to an evidence repository, AIR now explains how collected evidence is staged. If Direct Collection is off, the collection is written to the asset’s temporary evidence location before upload and requires sufficient free disk space. If Direct Collection is on, evidence is streamed to the repository with minimal asset disk usage.

The same information appears in task details through tooltips on Direct Collection rows. This helps analysts and administrators choose the appropriate collection mode for assets with limited disk space.

Investigation Hub Task Logs Show the Owning Asset

Task execution logs in Investigation Hub now include an Asset column. The asset name links through the existing asset-details redirect, so cloud assets open the Cloud Asset Info page and device assets open the device details page.

This improves investigation traceability by connecting log entries directly to the asset that produced them.

Tasking, Resource Management, and Responder Operations

Resource Limit Profiles

Resource Limits now use profile levels instead of exposing all numeric controls as the primary workflow. Analysts can select Minimum, Low, Balanced, High, Maximum, or Custom. The profile view shows CPU, CPU cores, memory, disk I/O priority, process priority, bandwidth, disk space reserve, and maximum execution time.

The Balanced profile is recommended by default. Users can preview a level before selecting it, and Custom mode shows how many values differ from the closest predefined level. This helps teams choose safe resource limits faster while retaining precise control when needed.

New task configurations and policy resource limits now default to Balanced, including a bandwidth limit and disk space reserve. Existing saved configurations keep their current values until changed.

Responder Uninstall After Organization Deletion

When an organization is deleted, AIR now creates the uninstall tombstones required for responders in that organization to receive signed uninstall tasks when they next check in.

AIR also keeps the organization-specific uninstall signing key so responders that already hold that key can validate the uninstall task. This prevents responders from remaining installed and repeatedly rejecting unsigned uninstall tasks after organization deletion.

One-Line Windows Deployment Command

The Windows Quick Deployment PowerShell command now copies as a single line with statements separated by semicolons. This improves compatibility with terminals and remote shells that collapse pasted newlines into spaces.

The command behavior is otherwise unchanged. The standard command still validates TLS, and the “without TLS verification” option remains available only where it was already offered.

Integrations

Brandefense TAXII Provider

Brandefense is now available as a predefined TAXII feed provider. Selecting Brandefense sets Basic authentication and pre-fills the discovery URL.

This reduces setup steps for teams that ingest threat intelligence from Brandefense into AIR.

Access Control Improvements

Organization and Case Scoping Refinements

AIR now rejects malformed organization IDs at the organization-scope guard boundary. Non-decimal inputs such as hexadecimal values, malformed bracketed values, decimal fragments, and comma lists on single-value fields return 400 instead of being coerced inconsistently.

Case name checks and duplicate-name suffixing are now scoped to the relevant organization and treat SQL wildcard characters literally. This prevents case name probing across organizations and keeps duplicate name handling tenant-scoped.

API v2 library writes now reject unknown organization IDs for all-organization callers, preventing objects from being bound to organizations that do not exist.

API Token Scoping and Audit Attribution

USER API tokens now respect private case membership. A token with a restricted role no longer inherits the issuing user’s global access to private cases.

interACT library upload and download paths now check the API token’s organization scope rather than the token owner’s organizations. Audit entries for API-token interACT library downloads now show the API token name and source API.

Multipart upload file IDs are now bound to the principal that created them, whether that principal is a user, API token, or responder. Other principals can no longer read, upload, finalize, or delete another principal’s multipart upload.

Authentication and Recovery Improvement

Password recovery now returns a uniform response for existing and non-existing accounts, preventing username enumeration through status-code differences.

Refresh token rotation now consumes the previous refresh token atomically and adds unique token IDs. Concurrent refresh attempts no longer mint multiple valid sessions from one refresh token.

OIDC SSO callbacks are now bound to the browser that started the sign-in through per-login state and PKCE. This prevents a callback from completing in a browser that did not initiate the login.

interACT shell APIs now enforce the configured interACT two-factor authentication requirement at the API layer, not only in the UI.

The Console now requires the JWT signing secret to be at least 20 characters at startup, matching the documented requirement.

WebSocket, Banner, and Confirmation Dialog Safety

WebSocket upgrades now validate the Origin header when it is present, while still allowing non-browser clients that do not send an Origin header.

Header banner markdown links now allow only safe site-relative and http or https targets. Unsafe link targets render as text or are rejected on save.

TAXII feed and Git repository confirmation dialogs now escape resource names before rendering confirmation copy, preventing stored markup from executing in another user’s session.

Evidence and Export Safety

Case notes CSV export now detects spreadsheet formula payloads even when they are preceded by whitespace, quotes, semicolons, or line breaks. AIR defangs these values before export.

Evidence repository create and update responses on the public API now return the same redacted DTO used by GET. Secret values such as passwords, access keys, and private keys are no longer returned in mutation responses.

Evidence PPC and named-file downloads now create audit events after successful downloads. Audit rows include caller context and organization scope without storing file bytes.

Settings

LOCARD Integration Removed

The legacy LOCARD blockchain integration has been removed from AIR Console settings and acquisition workflows as the upstream pilot project has concluded and is no longer actively maintained. Chain of Custody (LOCARD) settings, organization credential overrides, and the LOCARD option on acquisition tasks have been removed. Evidence is no longer submitted to a LOCARD account. RFC 3161 timestamping is unchanged. Tasks created before this change no longer return the stored LOCARD configuration when their task data is read.

Bug Fixes

  • Purge and Uninstall no longer fails on unmanaged assets with no security token. AIR now skips creating unverifiable uninstall tombstones when an asset has no responder security token, while still purging the Console record. Enrolled assets continue to get signed uninstall tombstones.

  • Dashboard task and asset counts now better match linked lists. The Tasks total aligns with View all, seven-day chart legends no longer navigate to mismatched all-time lists, asset segments include Off-network counts for users who can view assets, and Dashboard asset links apply the correct filters.

  • Assets no longer briefly show stale Dashboard filter results. When navigating from Dashboard links, AIR clears previous list state and guards against stale responses overwriting the new filtered results.

  • Processor proxy downloads can no longer render active content on the Console origin. AIR now restricts proxied response content types, forces attachment handling where appropriate, strips processor cookies, and adds defensive response headers.

  • Profile photo uploads now require real image data. AIR accepts only supported image data URIs whose decoded bytes match the declared image MIME type. HTML, SVG, and mislabeled data are rejected.

  • Header banner markdown links are now validated and sanitized. Unsafe link schemes are rejected on save and sanitized at render time, while safe http, https, and site-relative links continue to work.

  • TAXII and Git confirmation dialogs now escape names. Feed and repository names in cancel-sync and delete confirmations are escaped before rendering.

  • Investigation Hub evidence comments are sanitized. Unsafe HTML, images, styles, and script-capable markup are removed on save/read and before UI rendering.

  • Case-comment mention notifications now respect case access. Users who are mentioned but cannot read a private case no longer receive comment notifications for that case.

  • Recent Activity is now scoped to the signed-in user. AIR ignores foreign username filters and omits inaccessible private case data from recent activity responses.

  • Case Insights and statistics now respect private case visibility. Auto-close widgets, case summaries, live case counts, and utilization statistics now apply the same case access rules as the case list.

  • USER API tokens now respect private case membership. Tokens no longer inherit the issuing administrator’s private case access unless the token’s user is actually related to the case.

  • Case creation now blocks cross-organization owners. AIR verifies that an explicit case owner belongs to the case organization before creating the case.

  • Case name check and duplicate suffixing are now organization-scoped. The name availability check no longer searches all organizations, and wildcard characters in names are treated literally.

  • Organization ID parsing is now strict in authorization guards. Malformed values such as hexadecimal, unmatched brackets, decimal fragments, or trailing comma entries now return 400 instead of being coerced inconsistently.

  • Bulk LDAP sync is now scoped to the caller’s organizations. Organization-scoped users can no longer trigger directory synchronization for all organizations through the bulk sync route.

  • Bulk asset account sync is now scoped to the caller’s organizations. Organization-scoped users can no longer trigger synchronization of other organizations’ cloud asset accounts.

  • Cloud account read APIs now require the appropriate cloud account privilege. List, config, export, and account detail routes are no longer available to any authenticated user by default.

  • Organization detail responses no longer expose deployment tokens to view-only callers. The token is withheld unless the caller can deploy assets or manage deployment tokens.

  • User group reads no longer expose member security posture to every authenticated user. Fields such as login strategy, two-factor status, password state, and preferences are hidden unless the caller is allowed to view that information.

  • Trusted certificate and LDAP login settings now require Global Admin access. Users with settings-save permission alone can no longer change high-trust authentication or outbound trust settings.

  • Responder update policy filters are now validated. Organization override filters are checked against supported endpoint fields and safe values before storage and evaluation.

  • Triage custom destinations now validate evidence repository organization scope. Hunt/Triage tasks can no longer store another organization’s repository ID in save-to or send-to configuration.

  • Evidence repository public mutation responses no longer return decrypted credentials. Public create and update routes now return the same redacted contract as GET.

  • Multipart upload file IDs are owner-bound. A responder or user can no longer complete another principal’s pending multipart upload by knowing its file ID.

  • Off-network package reads no longer expose anonymous download links to non-creators. Organization peers can still read package metadata, but download URLs are withheld unless they are the creator or have all-organization access.

  • Legacy off-network archive password derivation was removed. Older generated-password routes now return 404. Current off-network archives use the configured archive password or no password, and legacy pre-4.28.0 archives that lack the modern metadata are rejected during import.

  • Webhook trigger tokens now require a stronger length range. V1 webhook create and update reject one-character, whitespace-only, too-short, and too-long tokens.

  • Webhook repeat-window enforcement is now concurrency-safe. Parallel receiver triggers with the same token now reserve the endpoint window atomically so only one task is created inside the minimum interval.

  • Webhook assignment polling is now scoped to integration-owned tasks. A webhook token can no longer read assignment metadata for arbitrary tasks it did not create.

  • Event subscription updates now invalidate removed event types. AIR clears webhook delivery cache entries for both previous and updated event types, reducing stale deliveries to old URLs after an event list is trimmed.

  • Git repository and event-subscription outbound test routes now require write-level privileges. View-only roles can no longer initiate caller-supplied outbound validation requests.

  • Git repository browse and branch routes now protect ad-hoc URLs. Requests that do not reference a stored repository now require create or update privileges.

  • Findings export validates sort fields before streaming. Invalid or unsupported sort fields now return 400 rather than failing during export generation.

  • Findings export filters unsafe object keys. AIR drops unsafe JSON-derived object keys before building export columns or data-structure metadata.

  • Finding verdicts and MITRE IDs are now validated. AIR rejects unsupported finding types and unknown MITRE tactic or technique IDs before persistence.

  • Investigation Hub finding writes no longer return 500 when optional fields are omitted. Missing optional path or MITRE list fields are treated as empty values.

  • Investigation Hub report writes now respect closed or archived cases. Report create and PDF generation are blocked on non-open cases with the same case-state rule used by other Hub writes.

  • Investigation update now requires manage-case. Updating shared investigation metadata, such as timezone, is no longer allowed with view-task alone.

  • Closed and archived cases now block Investigation Hub writes. Imports, comments, settings, findings, flags, exclusions, deletes, and related write paths now reject writes when the linked case is not open.

  • Task schedule updates now reject non-scheduled tasks cleanly. Updating a task that was created to run immediately returns 409 instead of 500. Invalid scheduler values such as unsupported execution modes are rejected with 400.

  • Fractional schedule repeat intervals are now rejected. Scheduled task configuration now requires whole-number repeat intervals.

  • Task statistics routes now require view-task and case scoping. Task aggregates now follow task visibility rules and case access narrowing for non-global administrators.

  • Refresh token rotation is now single-use and atomic. Concurrent reuse of the same refresh token no longer creates multiple valid sessions.

  • Forgot-password responses are now uniform. AIR no longer reveals whether a username exists through different status codes when password recovery is requested.

  • Two-factor verification for setup is now rate limited. The TFA verify route now uses the same per-IP throttle pattern as related authentication routes.

  • OIDC sign-in is now bound to the browser that initiated it. Per-login state and PKCE attempts prevent callbacks from completing in unrelated browser sessions.

  • WebSocket upgrades now validate Origin. Browser WebSocket requests must originate from the Console host or allowed development origins.

  • interACT API operations now enforce interACT two-factor prerequisites. Shell execution, cancel, close, and related routes now check the configured interACT 2FA rule before dispatching work.

  • interACT library upload and download now respect API token organization scope. Tokens can no longer act on library files outside their allowed organizations, and download audit rows now attribute the API token correctly.

  • PPC and named evidence file downloads now create audit records. Successful private, public, and v2 PPC downloads, plus named-file downloads, are now recorded with organization scope and caller attribution.

  • Processor proxy responses are now safe to open from the Console origin. AIR prevents processor-controlled HTML or script-capable responses from rendering as Console pages.

  • CSV exports now better protect against spreadsheet formula payloads. Whitespace-prefixed, quote-prefixed, semicolon-prefixed, and multiline formulas are neutralized.

  • Bulk purge-by-filter routes now reject unknown filter fields. Mistyped or unsupported filter keys no longer silently widen endpoint or task purge operations.

  • Bulk MITRE database version updates now validate organization scope. Filter organization IDs are checked before the bulk update handler runs.

  • Stale asset policies are now organization-scoped. Custom stale asset policies belong to one organization or all organizations, and cleanup evaluation can no longer claim assets across tenant boundaries unintentionally.

  • Update Settings now use the Management Service request model. Users can defer, schedule, cancel, and request updates immediately from the redesigned Update Settings page.

  • One-click interACT report links now work with strict cookies. AIR serves a same-origin transition page for interACT report destinations so newly set session cookies are sent correctly.

  • Windows deployment command copy is now single-line. The PowerShell command remains valid in shells that collapse newlines.

  • ESXi Collector delivery now uses the matching TACTICAL Cross version. This resolves customers receiving an outdated collector package through Console downloads.

  • Catalog-signed Windows binaries are handled more accurately in Cross-related evidence paths. Catalog signature handling improvements address cases where system binaries were previously reported as unsigned.

  • AppX catalog fallback no longer treats arbitrary files in package folders as signed. DRONE and TACTICAL Cross avoid reporting unrelated files as signed only because they sit in an AppX package folder.

  • macOS FSEvents collection is more responsive. The collection path now handles large FSEvents histories more efficiently, logs progress, and honors cancellation more reliably.

  • Disk-space checks no longer race under concurrent collection finalization. Internal disk-space state access is synchronized to avoid data races.

  • Azure repository upload errors no longer expose SAS signatures in responder logs. Credential-bearing query parameters are redacted from Azure error logging paths.

  • Responder TLS verification now rejects affected distrusted certificate chains across additional outbound clients. This strengthens TLS validation for responder communication and repository upload paths on builds that need compatibility checks.

  • interACT curl embedded trust store was refreshed. The embedded certificate bundle no longer trusts the affected roots and includes build checks to prevent reintroducing distrusted certificates.

  • Responder REST connections recover better from dead pooled sockets. Console REST requests can stop reusing a stalled connection after timeout and retry on a fresh connection.

  • Scheduled Tasks collection is being moved away from blocking legacy enumeration. The collector work reduces reliance on a live Task Scheduler service and supports more reliable completion and cancellation on affected Windows assets.

  • Scheduled Tasks evidence can detect hidden tasks more effectively. Raw registry TaskCache coverage supports detection of tasks hidden from Task Scheduler UI and COM enumeration.

  • DRONE and TACTICAL Cross signature checks now handle catalog-signed files more accurately. YARA whitelist decisions and executable enrichment are improved for Windows catalog-signed binaries.

  • Codex AI artifact parsing in TACTICAL Cross now captures more useful normalized events. VS Code Codex sessions now preserve message excerpts and tool actions more reliably while still retaining raw transcripts.