A digital identity security network links people, AI assistants, cloud services, and devices to a protected monitoring dashboard.
The week's identity and security announcements share one theme: identity teams now have to govern AI agents as well as people. Most of the items are vendor launches that arrived between November 16 and 18, and Solutions Review rounded them up on November 20. A few touch the Microsoft ecosystem directly: Semperis covers Entra ID, Pindrop has a Teams certification, Hexnode manages Windows PCs, and the roundup carries commentary on a Microsoft 365 outage.

Everything below is vendor-described. None of it has been independently tested, and nobody has confirmed the effectiveness claims.

A digital identity security network links people, AI assistants, cloud services, and devices to a protected monitoring dashboard. The Microsoft-facing items​

Semperis Lightning​

Semperis announced the release of Semperis Lightning on November 17. It is an identity resilience and security platform meant to help organizations prevent, detect, respond to and recover from agentic and conventional identity attacks. It is pitched as a single console for exposure management, attack-path analysis, human and non-human identity visibility, and detection-and-response across Active Directory, Microsoft Entra ID and Okta.

Here is what the announcement says it does:

  • Exposure scanning. A growing catalogue of more than 300 security indicators scans for exposures in human and non-human identities. It scores the results and provides one workflow for triaging them.
  • Inventory and attack paths. A single inventory of every identity, a focused view of attack paths leading to critical assets, and automatic discovery of non-human and agentic identities.
  • A live identity graph that maps relationships and attack paths, covering routes to domain administrators, privileged accounts and other critical assets.
  • Detection. Built-in rule templates and generative assistance flag suspicious Active Directory and Entra ID activity, including suspicious Kerberos and NTLM logon activity.

For Windows shops, the pitch is aimed at hybrid environments where Active Directory is still the backbone and Entra ID sits on top. Semperis also stresses recovery, meaning reversing risky directory changes and recovering Active Directory, Entra ID and Okta when prevention fails. Recovery is the harder half of identity security, and many programs are still building it.

The caveat is that the announcement doesn't publish independent validation. Pricing, licensing tiers and edition-level feature differences weren't in the material reviewed. If you run Entra ID, ask Semperis which of the Lightning capabilities apply to your tenant and licensing before you plan around it.

Pindrop and Microsoft Teams​

Pindrop announced Pindrop for Hiring on November 18. It supplies fraud signals across the hiring process to support investigation, and it doesn't make hiring decisions. The company also says its Pulse for Meetings integration with Microsoft Teams is now certified. Pindrop describes the Pulse integration as providing real-time synthetic audio and video detection, administrator controls, in-meeting alerts and connections to Microsoft security workflows. It says the integration is available through Microsoft Marketplace.

The certification is Pindrop's claim, and detection accuracy hasn't been independently tested. Teams admins will care mainly about deployment details, such as licensing and what meeting data the integration processes. Those weren't established here, so confirm them with the vendor.

Hexnode UEM​

Hexnode added native IT asset management to its UEM platform. It tracks enrolled endpoints, connected peripherals, bulk inventory and non-IT office assets from one console. The blog post lists several capabilities:

  • Equipment bundles assigned to users.
  • Checkout records with return deadlines.
  • Depreciation and book-value calculations.
  • Disposal certificates generated automatically.
  • QR-based audits.
  • A role limited to asset work.

The post also says asset management is available on Hexnode's Enterprise plan and above. That detail is from the Hexnode post, which wasn't searched in this pass.

The same post adds remote BIOS management for supported Windows PCs. It covers boot-order configuration and BIOS password controls, and the post says passwords can be set, changed or cleared on supported HP, Dell and Lenovo devices. Hexnode also added Shared Device Policies for Android. That is a practical item for admins who currently run vendor-specific scripts to manage firmware settings. As always, "supported" is the operative word, so check your hardware models.

The Microsoft 365 outage commentary​

The roundup also carries a commentary piece on the Microsoft 365 outage around Friday the 13th. Paul Robichaux of Keepit says the trouble began Thursday and ran well into Friday. He wrote that Microsoft hadn't yet published a root cause when he wrote the piece, though he expected one. Keepit sells SaaS backup, so the argument for independent copies of SaaS data suits its business. That doesn't make the point wrong. The commentary itself is the only evidence in the roundup, and no root cause is established in this material.

The AI-agent identity wave​

Linx Security​

Linx announced four capabilities on November 17:

  1. Shadow AI Detection. It pulls signals from EDR, MDM, CI/CD and cloud tools to find unsanctioned agents and map who uses them and what they can reach.
  2. Agent Runtime Control. Linx acts as the decision point for an AI gateway. It allows or blocks each tool call based on the identity behind it, the risk and the action. Linx's example is a finance agent that can read a CRM record but is blocked when it tries to export every customer record.
  3. Autopilots. Identity teams can build custom ones with defined scopes, human approval rules and dry runs before going live.
  4. Prompt-to-Workflow. A plain-language description becomes a tested governance workflow, such as cleaning up agents whose owner has left.

All four run on the Linx Identity Graph alongside human and non-human identities. Linx says 35% of its customers already secure agent identities with it, with more than 25,000 governed in production. Those are the company's own figures. Linx plans to demo the releases at the Gartner Identity & Access Management Summit in Las Vegas on December 7–9.

HYPR AgentPass​

HYPR's AgentPass ties AI agents to named human supervisors and evaluates requests before execution. It is described as spanning developer endpoints, a central agent gateway and real-time administrator controls. Listed features include passkey approvals, guardrails, continuous supervision and kill switches. I found no HYPR primary announcement in this pass, so treat the specifics and availability as the roundup's description only.

Brinqa​

Brinqa announced connected AI agents on November 17. Its press release describes:

  • Attack Path Insights, which maps potential routes to critical assets.
  • An Exploitability Agent that reassesses priorities as new evidence arrives.
  • A Validation Agent that tests supported exposures and retests them after remediation.
  • PlexTrac by Brinqa, where the Validation Agent retests supported CVEs that can be exploited remotely and records evidence on each finding.

Brinqa says recommendations combine customer environment data with threat intelligence. The word "supported" applies throughout, and Brinqa hasn't published coverage details for every exposure type.

Defense and response​

Keeper Security's emergency breach response​

Keeper announced on November 18 a global emergency breach response service for organizations under active attack. Submissions to an emergency portal go straight to its Solution Engineering triage team. Keeper says it will defer quotes and billing until the platform is deployed and access to sensitive resources is secured. The engineers help deploy KeeperPAM to revoke standing privileges, rotate compromised credentials, secure secrets and limit lateral movement.

This is a technical deployment path. It isn't a published incident-response guarantee, and no price, eligibility criteria or service-level commitment was in the release. Keeper's claim that cloud-native deployment takes minutes rather than months is also its own. A breached organization still needs a real incident-response process around any tool.

Humanix Complete Call Context​

Humanix announced this on November 17 and says it is available now to customers. Micro-agents compare sensitive service desk requests, such as password resets and MFA changes, with identity provider records, ITSM tickets and the organization's own procedures. Any contradiction or missing evidence produces an alert with a severity rating and a plain-language explanation.

Humanix gives an example. A caller claims to be an employee's manager and cites a ticket number that doesn't exist, while identity systems list someone else as the manager and show no prior verification. Service desk social engineering is a long-standing weak point in Entra and Active Directory environments, so correlating the call with identity and ticket records is a sensible idea. Detection performance hasn't been independently validated.

Core6 StorageGuard 10.1​

Core6 announced general availability of StorageGuard 10.1 on November 16. Its Change Control feature keeps a risk-scored timeline of configuration changes across scanned storage and backup systems. Teams can compare any system with an earlier point in time, mark changes as authorized or unauthorized, and trace findings to the changes that preceded them. AI-powered risk scoring is opt-in and runs on the customer's own approved AI provider. Existing customers get Change Control with the 10.1 upgrade.

Other additions in the release:

  • Remediation planning.
  • AI-drafted custom checks.
  • Guided onboarding.
  • An integration with Armis from ServiceNow.

The compatibility note matters for planning. The release notes say upgrades are available from 9.2.8 and above, and earlier installations must reach 9.2.8 or later first. They also warn that the upgrade can stop StorageGuard operations and take several hours in large environments. That detail comes from the research brief and wasn't re-checked here.

Ocean​

Ocean launched Agentic Phishing Simulation. It turns intercepted email threats from customer mailboxes into role-specific exercises and links Ocean's detection and response with employee readiness. I found no primary Ocean announcement. Confirm release status, data handling and supported mail environments before relying on it.

Governance, supply chain and data​

  • Fortra. On November 17 Fortra launched an AI module for GoAnywhere MFT and Globalscape EFT. The roundup lists natural-language project creation and modification, guided diagnostics and generated documentation for GoAnywhere. Globalscape EFT gets event diagnostics, rule analysis and reporting on operational trends and risk signals. Availability tiers and prerequisites weren't established.
  • OneTrust. The Fall '26 release spans AI governance, privacy operations, consent management and third-party risk. The roundup mentions a foundation-model intelligence library, expanded dependency visualization, and dashboards covering usage, risk, evaluation coverage, potential personal-data exposure and token consumption. It also adds protections against abuse of automated data-subject request processes. No primary release was verified.
  • Ory. Ory is doubling the frequency of free security updates for open-source users. It is also introducing a Continuous Security License, a cheaper alternative to a full Enterprise License for organizations needing more frequent vulnerability fixes. The exact cadence and license terms weren't verified, so don't infer patch SLAs from the summary.
  • Spur Intelligence. Its AI Data Feed covers about 150,000 IP addresses across a growing number of operators. It is meant to help teams identify AI services, including validated ones operating outside their published IP ranges. Methodology and update frequency weren't verified.

What admins should take from it​

Three practical points stand out.

  1. Inventory agents like any other identity. Several launches assume you need an owner, an access map and a kill switch for each AI agent. That holds regardless of which vendor you buy from.
  2. Don't skip the service desk. Voice attacks on password and MFA resets exploit process, not software, so write the verification procedure down before buying anything.
  3. Check the fine print. Licensing tiers, supported platforms and upgrade prerequisites are where vendor announcements get vague. The Core6 downtime warning and the Hexnode plan tier are examples of details that only show up on a second read.

Treat this as a map of where vendors are heading, not a buying guide. Nothing here replaces a trial in your own environment.

 

References

  1. Identity Management and Information Security News for the Week of November 20th: Linx Security, Semperis, Keeper Security, and More - Solutions Review Solutions Review Fri, 20 Nov 2026 16:29:42 GMT
  2. Linx Launches Shadow AI Detection and Agent Runtime Control linx.security
  3. Semperis Unveils Lightning Platform to Accelerate AI Identity Detection and Response - Semperis semperis.com