This piece sets out what each vendor says it does, where Microsoft's own documentation draws the boundaries, and which claims are vendor assertions that nobody has independently confirmed.
Three tools, three moments in the lifecycle
| Stage | Vendor | What it says it does |
|---|---|---|
| Screening and interviews | Pindrop | Gathers fraud signals from résumé review and live interviews, and adds deepfake detection in Teams |
| Onboarding (joiner) | 1Kosmos | Adds identity proofing to SailPoint's joiner process |
| Later high-risk moments | IdRamp | Issues reusable Entra Verified ID credentials for help desk, interviews and executive calls |
They are complementary rather than interchangeable. One looks at candidates, one gates access at onboarding, and one gives an employee something to present later.
Pindrop: hiring signals plus Teams deepfake detection
Pindrop launched Pindrop for Hiring on November 18, and in its own announcement the company describes it as a product that brings fraud signals together from résumé review, virtual interviews and investigation workflows inside enterprise hiring platforms. AI-based résumé review flags fabricated employment histories, suspicious identities and duplicate applications. During interviews, Pulse for Meetings analyzes live audio and video for signs of synthetic media. Voice analysis and location-risk signals add more material for investigators.
Pindrop is explicit that the product supplies risk signals to support a company's own fraud investigation and identity processes, and that it does not make hiring decisions. That distinction matters. A flag is a lead for a human, not a verdict on a candidate.
The Teams certification
For Microsoft shops, the more relevant news is that Pindrop says its Pulse for Meetings integration with Teams is now certified. According to the company:
- Teams administrators can manage detection through existing Teams controls.
- Meeting participants see warnings when potential synthetic media is detected.
- Relevant detection events can flow into Defender security workflows.
- The integration is available through Microsoft Marketplace.
Albert Kooiman, Microsoft's general manager for Teams Partner Engineering and Customer Enablement, is quoted saying the integration gives customers real-time synthetic-media signals for their broader security operations. Certification is not a performance guarantee. It tells you the integration meets Microsoft's Teams app requirements, not how accurate the detection is. Treat accuracy as something to test in your own pilot.
The numbers, and why to hold them loosely
Pindrop reports the following from its own data:
- One in six applicants it analyzed showed signs of fraud.
- Applicants with indicators of potential North Korean links rose about 630 percent year over year, from one in 343 in 2025 to one in 47 in 2026.
- One in four of those applicants used deepfakes in interviews.
- Its 2026 Deepfake Readiness Index found 74 percent of surveyed security leaders had encountered or suspected deepfake attacks, while 10 percent had purpose-built detection tools.
These are company-reported figures from a company that sells detection. They are not independently established prevalence rates, and the sample is not necessarily representative of all employers. The trend is plausible and consistent with the threat the Justice Department has described. The precise ratios should be read as one vendor's view.
Pindrop also tells an anecdote about its own recruiting. An applicant's email address appeared in a Justice Department disclosure about North Korean IT-worker schemes, and Pulse for Meetings flagged an IP location inconsistent with the stated Texas home. That is a vendor account, not a published independent investigation.
One detail from the original report needs care. It says Christina Marie Chapman was sentenced in July 2025 to 102 months. I could not confirm that sentence in the sources available for this piece, so I am not repeating it. What the Justice Department's account of the scheme does support is that she helped North Korean IT workers get remote jobs, and that U.S.-based facilitators hosted employer-issued laptops so overseas workers appeared local. The scale is more than 300 U.S. companies and about $17 million in proceeds. That is the pattern these products are aimed at.
1Kosmos: proof of identity before access
1Kosmos says it has joined SailPoint's Technology Alliance Partner ecosystem and integrated with SailPoint Human Fabric. The aim is to add identity verification to the joiner process, so an organization can confirm a new identity before making access decisions.
The checks combine document-based proofing, biometric verification and liveness detection, with optional matching of identity attributes against data already in SailPoint. 1Kosmos says sessions typically take under two minutes. That is a vendor estimate, not a guarantee. The company says the same capability extends to account recovery, password resets and access provisioning.
How the setup works
1Kosmos's own integration documentation gives a useful picture of what administrators would actually do. It describes a SailPoint Identity Security Cloud setup using Workflow Builder, and it may reflect the current implementation rather than the exact launch configuration:
- Create a source in SailPoint to track identity-verification status.
- Build an enrollment workflow that fires when a new identity is created. An HTTP request calls the 1Kosmos API, which returns a session ID and verification URL. Status is recorded as PENDING and the user is emailed the link.
- Build a second workflow that receives the completion callback, retrieves the result summary and sets the status to COMPLETED.
- In the 1Kosmos admin portal, choose document types, biometric checks and liveness settings, and optionally enable matching of first name, last name and date of birth against SailPoint attributes.
- Test both workflows with a test identity.
- Wire the status into provisioning rules, certification campaigns or manual review queues.
Prerequisites include an active SailPoint Identity Security Cloud tenant with enough permissions to create sources and workflows, a 1Kosmos tenant, and API credentials. The documentation says it uses standard HTTP Request nodes and no custom code. All verification logic runs in 1Kosmos, and SailPoint acts on the outcome through existing governance workflows.
The takeaway is that this produces verification evidence that feeds governance. It is not an automatic access decision. What a "failed" or "pending" status actually blocks is up to your own provisioning and review rules.
IdRamp: a reusable credential on Entra Verified ID
IdRamp's Verified ID Studio is the piece most squarely in Microsoft territory. Per Biometric Update, it is an administrator console for designing, issuing and tracking Microsoft Entra Verified ID credentials for a workforce, replacing hand-written credential definitions and a custom issuing app.
The reported flow:
- The employee is verified through an identity verification provider the organization chooses, using a government ID and live facial capture, and the verified face image becomes the credential's photo.
- The employee holds the credential in Microsoft Authenticator.
- It can be presented at risky moments such as help desk MFA resets, remote interviews and video calls with executives.
- Microsoft Entra Face Check can then compare the person's live face with the verified image.
- Administrators can issue by email, link or QR code, revoke and replace credentials, and write the verification back to the Entra account.
- Credentials are issued under the organization's own Verified ID authority and domain, not IdRamp's.
A caveat on sourcing: I found no independent or official IdRamp announcement confirming the Studio product or its November 17 date, so these specifics rest on Biometric Update's report. IdRamp does have a documented Microsoft relationship. It holds Microsoft Security Elite Partner status with an Entra Verified ID specialization, and its Microsoft Marketplace listing for Verified Workforce says it binds every Microsoft Entra ID account to a verified human identity.
What Microsoft says Verified ID is, and isn't
Microsoft's documentation sets limits that any admin evaluating this should know:
- Verified ID provides cryptographic proof of a user's verified identity, but it cannot satisfy authentication requirements such as sign-in, MFA or self-service password reset. It is an identity-proofing layer for situations where identity must be re-established, such as account recovery.
- Face Check compares the photo on the user's credential or ID with a real-time biometric check to confirm presence. It requires a Face Check license, available through Entra Suite or standalone.
- Identity verification profiles require Microsoft Entra ID P1 or an eligible higher license, and during preview are available only in the Microsoft public cloud, not national or sovereign clouds.
- Profiles can run in Evaluation or Production mode. A successful evaluation does not complete onboarding or recovery or grant access.
- Profile changes can take 5 to 10 minutes to apply.
- A sign-in error, AADSTS10020005 ("No valid VidProfile data found for the tenant"), can appear when no valid profile is available to a user. Microsoft says to check the profile, the user's group assignments and exclusions, and whether Verified ID is enabled.
- For resource-access verification under Conditional Access, Microsoft says to select Browser and Mobile apps and desktop clients, because legacy client types can make the flow fail.
In practice, a tool like Studio does not replace Entra's native controls. It helps issue and manage the credentials that Entra's verification features consume. Licensing and tenant configuration will still determine what you can actually deploy.
A crowded field
Biometric Update notes that others are working the same ground:
- BrightHire, a Zoom subsidiary, flags deepfakes, IP location and AI-generated answers in live interviews.
- Checkr's Fraud Detection Suite pairs Socure-powered biometrics with document, liveness and deepfake checks.
- Deel plans to acquire Clarity to add identity verification and deepfake detection to Deel IT.
- First Advantage and ID.me have partnered from recruitment through onboarding.
- World ID for Zoom verifies participants with a one-time Orb scan and a Human Badge on video tiles.
What IT and security teams should take from this
- Map the lifecycle first. Decide where identity is actually confirmed today: candidate, new hire, help desk, executive call. Most programs verify once at the front door and assume everywhere else.
- Keep humans in the loop. Pindrop says outright that its signals don't make hiring decisions, and 1Kosmos feeds results into review queues. Define who acts on a flag.
- Check licensing early. Face Check and identity verification profiles carry Microsoft license requirements separate from any partner's fee.
- Pilot before trusting accuracy claims. Detection rates, verification times and fraud statistics here are vendor-reported.
- Mind privacy and biometrics law. Biometric collection from candidates and employees raises consent and retention questions. Microsoft advises reviewing a provider's privacy, data retention and compliance policies before moving profiles to production.
None of these tools alone guarantees someone is who they claim to be. They reduce specific gaps, and they work best alongside existing access policy and human judgment. The harder question isn't whether a vendor can spot a deepfake in a Teams call, but whether your help desk will actually stop and ask for proof when the caller sounds like the CEO.
References
- Pindrop, 1Kosmos and IdRamp target workforce identity gaps from hiring to help desk - Biometric Update Biometric Update · 2026-11-19T18:50:31+00:00
- Pindrop Launches Pindrop for Hiring; Pulse for Meetings Integration Earns Microsoft Teams Certification pindrop.com
- Verified Workforce by IdRamp marketplace.microsoft.com