New Zoom Bug Prompts Security Fix, Platform Changes
A newly discovered Zoom vulnerability would have enabled an attacker to join active meetings and access audio, video, and documents shared.
CPX 360 – New Orleans, La. – A previously undisclosed and now patched vulnerability in the Zoom conferencing platform could have let attackers drop into active meetings by generating and verifying Zoom IDs.
Zoom users know the platform’s unique meeting IDs are made up of 9, 10, or 11-digit numbers. If hosts don’t require a conference password or enable the Waiting Room feature, Zoom ID is the only factor protecting meetings from unauthorized attendees. Check Point researchers found it was possible for an attacker to generate potentially valid Zoom IDs and automate their verification.
“The number should be privately shared, and it should be that nobody should be able to guess it,” says Check Point head of cyber research Yaniv Balmas. “We found a vulnerability in Zoom that allows it to tell us whether a number is a meeting number in a matter of minutes.”
Researchers pre-generated a list of potential meeting IDs and prepared a URL string for joining a meeting. When the URL was entered with a random meeting ID number, they noticed the HTML body of the returned response indicated “Invalid meeting ID” or “Valid Meeting ID found,” depending on whether the ID was linked to an active conference. Automating this approach allowed them to quickly determine valid ID numbers and drop in on random ongoing calls.
“Although we know the number is valid, we don’t know whose chat it’s going to be,” Balmas notes. “You can call it Zoom roulette.”
Exploiting this vulnerability could grant an attacker the same privileges as any Zoom attendee, meaning they would have access to audio, video, and documents shared during the call. While the intruder would not be invisible, Balmas points out, it wouldn’t be difficult to go unnoticed.
“To be frank, when you have a meeting with 20 to 30 participants, do you take the time to validate each and every participant in the meeting?” he asks. Researchers were able to correctly predict roughly four percent of the randomly generated meeting IDs, which they consider a high chance of success compared with pure brute force.
This isn’t a simple attack, Balmas says, but an intermediate adversary could pull it off. Other vulnerabilities the research team uncovers are typically more technical, he explains. “I wouldn’t call this one easy, but the bar is definitely lower than what we usually do,” he adds.
The Check Point Research team discovered this flaw last year and contacted Zoom in July 2019. Following a responsible disclosure process, the communications company released a fix in August of 2019 and introduced several mitigations to its platform, so this type of attack is no longer possible. The patch must be deployed manually.
Zoom is adding passwords by default to all future scheduled meetings. Users can add a password to meetings they have scheduled; Zoom is sending instructions to users. Password settings can be enforced at the account and group levels by the account administrator.
Zoom no longer automatically indicates if a meeting ID is valid or invalid; this way, an attacker will not be able to narrow the pool of meetings in an attempt to join one. Repeated attempts to scan for IDs will cause a device to be blocked for a period of time.
Who’s On the Line? Poking Holes in Conference Platforms
This is the latest in a series of vulnerabilities discovered in popular conference platforms. Late last year, the CQ Prime Threat Research Team disclosed the “Prying-Eye” flaw, which existed in the Zoom and Cisco Webex conferencing tools. Prying-Eye could let attackers scan for, and drop into, video meetings unprotected by a password.
More recently, Cisco issued a patch for CVE-2020-3142, a vulnerability in Cisco Webex Meetings Suite websites and Cisco Webex Meetings Online that would allow an unauthenticated, remote attendee to join a password-protected meeting without entering the password. An attacker could exploit this by accessing a known meeting ID or URL from a mobile device’s Web browser.
“Video chats are a very valid attack vector,” says Balmas. “I wouldn’t be surprised if [attackers] are trying to look for more vulnerabilities there.” Flaws like these could grant intruders access to directors’ meetings and other calls where participants discuss sensitive business matters.
Kelly Sheridan is the Staff Editor at Dark Reading, where she focuses on cybersecurity news and analysis. She is a business technology journalist who previously reported for InformationWeek, where she covered Microsoft, and Insurance & Technology, where she covered financial … View Full Bio