Skip to main content
Today: Today February 19, 2026
HubNews
Blockchain+
Cybersecurity+
Development+
Economy & Finance+
Gaming+
Artificial Intelligence+
Hardware+
Startups
Blockchain+
Cybersecurity+
Development+
Economy & Finance+
Gaming+
Artificial Intelligence+
Hardware+
Startups

HubNews

Receive weekly the main news and analyses about Artificial Intelligence directly in your email.

Sign Up for Free

News

  • Home Page
  • Feed
  • Guides
  • AI Products
  • Top
  • Deep Dives
  • Search

More

  • Games
  • Tools
  • Subscribe Free
  • Podcast

Information

  • About Us
  • Contact
  • FAQ
  • Developers
  • Sponsors

Legal

  • Privacy Policy
  • Terms of Service

© 2026 HubNews.ai. All rights reserved.

Cybersecurity
Amazon fixes flaw that threatened 66% of cloud infrastructure

Amazon fixes flaw that threatened 66% of cloud infrastructure

TL;DR

A critical flaw in Amazon Web Services (<strong>AWS</strong>) <strong>CodeBuild</strong> service posed a risk to 66% of cloud computing environments globally. Identified by security firm <strong>Wiz</strong> and dubbed <strong>CodeBreach</strong>, the vulnerability was patched in September 2024 before malicious actions could occur.

www.tecmundo.com.br•January 15, 2026•
3 min read
•0 views

A critical flaw in the CodeBuild service of Amazon Web Services (AWS) posed a risk to 66% of global cloud computing environments. Identified by the security firm Wiz and called CodeBreach, the vulnerability was patched in September 2024, before malicious actions could occur.

The vulnerability originated from regular expressions used to validate user IDs in CodeBuild. The absence of special characters ^ (start) and $ (end) allowed improper IDs to be accepted, enhancing the risk of a supply chain attack.

How the vulnerability was identified

The Wiz team investigated AWS’s continuous integration pipeline after noticing an attack on the Amazon Q extension for VS Code. During their analysis, they discovered that the access control filter of CodeBuild was misconfigured.

The filter known as ACTOR_ID operated with a permissions list that was not restrictive enough due to flaws in the regular expressions. This allowed any ID containing a trusted ID to pass the security check.

Ease of exploitation

The team demonstrated the ease of exploiting the flaw by creating 200 automated applications on GitHub, generating sequential IDs. In this way, a malicious ID was quickly identified that managed to bypass the security filters.

The researchers then proposed an apparently legitimate commit with hidden code to steal GitHub credentials during compilation, which could have disastrous results.

The extent of the risk

The flaw affected significant repositories of AWS, with the most critical being the AWS SDK for JavaScript, used in 66% of cloud environments. Its issues could have serious consequences, especially due to its presence in the AWS management console.

Comparison with previous incidents

The potential impact was compared to the SolarWinds attack in 2020, which compromised around 18,000 customers. In this context, the possibility of a similar supply chain attack on AWS raised significant concerns, as it could grant direct access to sensitive information and critical systems.

Hypothetical attack scenario

If criminals had discovered the vulnerability, they could have created several applications to obtain a malicious ID, passing security checks and inserting malicious code in a pull request. Compromising the SDK could lead to the installation of a backdoor, with severe consequences for millions of applications.

Rapid resolution by AWS

After being notified by Wiz in August 2024, AWS fixed the flaw within 48 hours, adjusting the regular expressions to include the necessary anchors. Audits ensured that no exploitation occurred by other agents, and additional security measures were implemented to prevent future incidents.

A thorough analysis of CloudTrail logs reaffirmed the certainty that system security had been restored and new protection methods were implemented in the build processes.

Content selected and edited with AI assistance. Original sources referenced above.

Share

Sources

www.tecmundo.com.br

Primary
https://www.tecmundo.com.br/seguranca/409897-amazon-corrige-falha-critica-que-poderia-ter-comprometido-66-dos-ambientes-cloud-do-mundo.htm

Jan 15, 2026

Enjoyed this article?

Get the best tech news delivered to your inbox every day.

Comments

Write a comment

More in Cybersecurity

Criminals Steal Passwords from Brazilians with Fake Apple Websites
Cybersecurity

Criminals Steal Passwords from Brazilians with Fake Apple Websites

A phishing scheme targets Brazilian iPhone users, using over <strong40 fraudulent domains</strongto collect Apple ID credentials.

HubNews • FEB 11 • 1 min read
Phishing Scam Uses Leaked CPF to Steal Money via PIX
Cybersecurity

Phishing Scam Uses Leaked CPF to Steal Money via PIX

A new phishing campaign exploits leaked CPF data in Brazil to steal money through PIX, the instant payment system, by sending fraudulent SMS messages.

HubNews • FEB 9 • 1 min read
Hackers Exploring OpenClaw for Malware Distribution
Cybersecurity

Hackers Exploring OpenClaw for Malware Distribution

Researchers from Bitdefender Labs have identified that OpenClaw, an automation platform utilizing artificial intelligence, is being used as a means for malware distribution. Analysis revealed that among the skills examined, approximately 17% exhibited malicious behavior in February 2026.

HubNews • FEB 7 • 1 min read