Using Teams' Guest Access feature? You're opening a security hole

Microsoft Teams’ Guest Access feature can leave users exposed if they connect to an insecure environment, and attackers can now invite them at scale.

Microsoft Teams has fast become the go-to collaboration platform for businesses, and its ability to welcome external users to collaborate and chat in real time is fantastic for productivity.

But a new report from security firm Ontinue warns that connecting to a clients’ external environment is not always the safest option, no matter the efficiency gains.

Threat researcher Rhys Downing says , “Effective protection is not inherent to [Teams]; it emerges from how each tenant chooses to manage external access, identity boundaries, and integrated security controls.”

The security gap doesn’t come from a flaw in Teams, but simply "the reality of cross-tenant collaboration.” That means users operating as guests in another environment aren’t protected by the same policies and processes as their own tenant.

To put it another way: attackers can bypass all Defender for Office 365 protections, like Safe Links, zero-hour auto purge and malware scanning, by inviting users into malicious tenants.

Why is Ontinue highlighting this now? Because Microsoft introduced a new feature this month that makes it easier for attackers to bring victims into their own environments.

That feature is catchily known as MC1182004 , and it allows Teams users to launch a chat with anyone with an email address. Recipients receive an invitation through Teams or by email (from Microsoft, so not normally picked up by email security) to join as a guest.

Importantly, the ability to launch these invitations is enabled by default, but security is not on low-cost licenses (Teams Essentials, Business Basic or trial). That means attackers can cheaply and easily set up insecure tenants, then invite victims en masse.

Security teams can disable the ability to send the invites with a PowerShell command , but there is no way to protect against inbound invitations.

Ontinue’s Rhys calls the feature “a perfect attack vector for threat actors who understand how cross-tenant security actually works.”

Once a user is in the insecure environment, the attacker has a variety of options. They can send phishing links without detection; distribute malware without alerts; conduct social engineering; and exfiltrate data under false pretences.

There are ways to protect your users. Ontinue recommends: