Security

Patches

You probably want to patch this critical GitHub Enterprise Server bug now

Unless you're cool with an unauthorized criminal enjoying admin privileges to comb through your code


A critical bug in GitHub Enterprise Server could allow an attacker to gain unauthorized access to a user account with administrator privileges and then wreak havoc on an organization's code repositories.

The good news is that there's a fix. The Microsoft-owned code hosting service addressed the 9.5 CVSS-rated flaw tracked as CVE-2024-6800 in GitHub Enterprise Server (GHES) versions 3.13.3, 3.10.16, 3.11.14, and 3.12.8.

Orgs running a vulnerable instance of GitHub Enterprise Server (GHES), GitHub's self-hosted version, will likely do well to download the update ASAP as miscreants are likely already scanning for this CVE.

Affected versions of GHES include 3.13.0 to 3.13.2, 3.10.0 to 3.10.15, 3.11.0 to 3.11.13 and 3.12.0 to 3.12.7.

As GitHub explained in the release notes we’ve linked to above, the critical flaw affected GHES instances that use Security Assertion Markup Language (SAML) for single sign-on authentication. The SAML authentication allows specific identity providers (IdPs) that use publicly exposed and signed federation metadata XML. This could allow an attacker to forge a SAML response to gain administrator privileges on a compromised machine, thus giving an unauthorized party access to your organization's GitHub-hosted repos.

This vulnerability, along with two others addressed in version 3.13.3, were reported via the GitHub Bug Bounty program.

The other two now-fixed flaws are both rated medium-severity.

CVE-2024-7711 could allow an attacker to update the title, assignees and labels of any issue inside a public repository — public being the key word here. Private and internal repositories are not affected by this bug, which earned a 5.3 CVSS rating.

CVE-2024-6337 is a 5.9-rated vulnerability that could allow an attacker to disclose the issue contents from a private repository using a GitHub App with only 'content: read' and 'pull_request_write: write' permissions.

This one can only be exploited with a user-access token, we're told. Installation access tokens are not affected.

It's been a rocky couple of weeks for the collaborative coding colossus.

This security update comes about a week after GitHub broke itself after rolling out an "erroneous" configuration change to all GitHub.com databases. This caused a global outage to several of its services, along with GitHub.com and the GitHub API.

Also last week, Palo Alto’s Unit 42 threat intelligence team found that a bad combination of misconfigurations and security flaws can make GitHub Actions artifacts leak both GitHub and third-party cloud services tokens. ®

Send us news
Post a comment

Mitel 0-day, 5-year-old Oracle RCE bug under active exploit

3 CVEs added to CISA's catalog

GoDaddy slapped with wet lettuce for years of lax security and 'several major breaches'

Watchdog alleged it had no SIEM or MFA, orders rapid adoption of basic infosec tools

Biden signs sweeping cybersecurity order, just in time for Trump to gut it

Ransomware, AI, secure software, digital IDs – there's something for everyone in the presidential directive

Microsoft eggheads say AI can never be made secure – after testing Redmond's own products

If you want a picture of the future, imagine your infosec team stamping on software forever

Miscreants 'mass exploited' Fortinet firewalls, 'highly probable' zero-day used

Ransomware 'not off the table,' Arctic Wolf threat hunter tells El Reg

CISA: Wow, that election had a lot of foreign trolling. Trump's Homeland Sec pick: And that's none of your concern

Cyber agency too 'far off mission,' says incoming boss Kristi Noem

Look for the label: White House rolls out 'Cyber Trust Mark' for smart devices

Beware the IoT that doesn’t get a security tag

Ransomware crew abuses AWS native encryption, sets data-destruct timer for 7 days

'Codefinger' crims on the hunt for compromised keys

OpenAI's ChatGPT crawler can be tricked into DDoSing sites, answering your queries

The S in LLM stands for Security

FCC to telcos: By law you must secure your networks from foreign spies. Get on it

Plus: Uncle Sam is cross with this one Chinese biz over Salt Typhoon mega-snooping

Just as your LLM once again goes off the rails, Cisco, Nvidia are at the door smiling

Some of you have apparently already botched chatbots or allowed ‘shadow AI’ to creep in

After China's Salt Typhoon, the reconstruction starts now

If 40 years of faulty building gets blown down, don’t rebuild with the rubble