Criminals are using fake app invitations and support messages to trick people into granting access to online accounts. This attack, known as OAuth consent phishing, can let an attacker read email, view files, or send messages without first stealing the victim’s password.
How consent phishing works
OAuth is a standard that lets one application request access to another service. For example, a calendar app might ask to view a user’s schedule. That request can be legitimate, but attackers can create applications with familiar names and promote them through phishing messages.
The FBI says recent campaigns have sent malicious links through commercial messaging services while impersonating government officials, media representatives, or other publicly known people. The link opens an authorization page that appears legitimate. If the user selects “Accept,” the attacker’s application can receive permission to access the approved data. The FBI calls this technique “OAuth consent phishing.” The FBI’s September 2026 advisory says the activity has targeted prominent victims, their family members, and personal acquaintances since late 2025. The advisory also describes earlier campaigns that impersonated event coordinators and planners.
Permissions that deserve caution
Some requests create more risk than others. Be especially careful when an application asks for access to email, files, contacts, mailbox settings, or the ability to send messages. Microsoft says attackers commonly misuse combinations of mail, contacts, mailbox settings, people, files, and notes permissions in consent-phishing attacks. Its incident-response guidance says attackers used a combination of those first six permission groups in 99% of the consent-phishing attacks observed by Microsoft’s incident response team. Microsoft’s application-consent guidance explains what those permissions can expose.
Before approving a request, examine the application name, publisher, website domain, and exact information it wants to access. A familiar logo does not prove that an application is genuine. Microsoft warns that attackers can copy popular product names or disguise an app as a trusted service. Broad access is unusual for a simple meeting, document, or event invitation.
Safer approval decisions
- Pause before approving. Treat an unexpected permission prompt as a security decision, not a routine login. If the request came from an email, text, social-media message, or phone call, open the service through its known website or app instead of using the supplied link.
- Read every permission. Look for access to email, cloud files, contacts, or the ability to send messages. If the request does not match the task, select Cancel.
- Check the publisher and domain. Compare the application’s publisher and website with information from the company’s official site. Small spelling or grammar errors, unrelated domains, and vague publisher names are warning signs.
- Prefer verified applications. Microsoft recommends using publisher verification and restricting consent to applications that an organization trusts. Personal users should favor apps that they intentionally installed and understand.
- Do not share verification codes. A caller or message that asks for a one-time code is not helping you approve an app. The Federal Trade Commission says anyone asking for an account verification code is a scammer. The FTC’s consumer guidance says to contact the company through a trusted number or official app instead.
If you approved a bad app
Act quickly, but do not continue communicating with the person who sent the request. Open the account provider’s official security page and review connected applications, authorized applications, or third-party access. Revoke access for anything unfamiliar. The FBI warns that changing a password alone does not revoke an OAuth token. Microsoft says an organization should disable a malicious application so it cannot obtain new tokens or be approved again. Simply deleting the application may not prevent it from returning. Microsoft’s response playbook describes this process for Microsoft Entra environments.
If you entered a password on a suspicious page, change it through the official account site and review active sessions and recent sign-ins. Check for unfamiliar forwarding rules, sent messages, new authentication methods, or changed recovery details. If the account is used for work, school, banking, or health services, notify the organization’s security or support team through a trusted channel. Also consider whether the attacker could have viewed files or messages containing identity documents, financial information, or other sensitive data.
Why this threat is growing
Consent phishing is attractive because it abuses a normal cloud authorization feature instead of relying on a software vulnerability or directly stealing a password. Microsoft explains that the attacker’s application can exchange an authorization code for an access token and use that token to make API calls on the user’s behalf. The FBI warns that this access can persist until the victim revokes the authorization in the application’s security settings.
That risk can extend beyond the original approval. Microsoft has reported attacks that combine social engineering, passkey-themed lures, and device-code phishing to compromise cloud identities and then collect files or email. Microsoft’s September 2026 research describes this broader pattern.
The safest habit is simple: treat an unexpected app-approval screen as a security decision, not a routine button click. When the request is unexpected, asks for broad access, or comes from an unverified contact, stop and verify it through a separate trusted channel.
References
Federal Bureau of Investigation. (2026, September 1). Malicious cyber actors gain access to victim accounts through consent phishing. Internet Crime Complaint Center. https://www.ic3.gov/PSA/2026/PSA260901
Federal Trade Commission. (2024, March 7). What’s a verification code and why would someone ask me for it? Consumer Advice. https://consumer.ftc.gov/consumer-alerts/2024/03/whats-verification-code-why-would-someone-ask-me-it
Microsoft. (n.d.). App consent grant investigation. Microsoft Learn. Retrieved November 20, 2026, from https://learn.microsoft.com/en-us/security/operations/incident-response-playbook-app-consent
Microsoft. (n.d.). Compromised and malicious applications investigation. Microsoft Learn. Retrieved November 20, 2026, from https://learn.microsoft.com/en-us/security/zero-trust/security-operations-playbook-app-consent
Microsoft Security Research. (2026, September 9). Passkey-themed social engineering leads to identity and cloud compromise. Microsoft Security Blog. https://www.microsoft.com/en-us/security/blog/2026/09/09/passkey-themed-social-engineering-leads-identity-cloud-compromise/
