All security news

Daily brief

Binance's Record Heist, Ivanti Zero-Days, and Active GitLab Exploits

Written automatically by AI from the day's headlines. Check the linked sources before you act on it.

North Korean Hackers Linked to Historic Binance Crypto Theft

The FBI and U.S. Secret Service have officially blamed North Korean state-sponsored hackers known as TraderTraitor for stealing over $6 billion from cryptocurrency exchange Binance. Federal authorities have flagged more than 600 wallet addresses linked to the incident as the group attempts to launder the proceeds. Crypto platforms and organizations handling digital assets are urged to monitor and block transactions associated with these addresses.

Sources: SecurityWeek, The Record, BleepingComputer, The Hacker News

Ivanti Patches Connect Secure Zero-Days Under Active Attack

Ivanti has issued urgent patches for two Connect Secure remote access gateway vulnerabilities that attackers have chained together in zero-day intrusions since at least October 1. The Cybersecurity and Infrastructure Security Agency added both flaws to its Known Exploited Vulnerabilities catalog following confirmed attacks. Network administrators should apply Ivanti’s latest updates without delay to protect exposed appliances.

Sources: BleepingComputer, SecurityWeek

Hundreds of Exposed GitLab Servers Targeted in Active Exploits

More than 1,000 internet-facing GitLab servers remain vulnerable to a maximum-severity flaw currently being targeted in remote code execution attacks. The ongoing wave of attacks recently led the Alpine Linux team to rebuild its development server and issue blanket password resets after detecting an attempted breach. System administrators running self-hosted GitLab installations should patch immediately and review access logs for unauthorized activity.

Sources: BleepingComputer, BleepingComputer

Polymarket Traders Targeted Through Hijacked Third-Party Script

Attackers manipulated domain registrar settings for performance monitoring service DebugBear to deliver malicious code directly to prediction market Polymarket. The injected script automatically liquidated positions and redirected millions of dollars from signed-in traders before the platform refunded affected users. Developers should tightly audit third-party script dependencies and restrict external code execution through content security policies.

Sources: SecurityWeek, The Hacker News

Critical Flaws in Langflow AI Tool Allow Remote Code Execution

IBM disclosed two critical security vulnerabilities in Langflow, a widely used open-source tool designed for constructing artificial intelligence agents and workflows. The flaws permit remote attackers with no user accounts to bypass access controls or execute arbitrary code on the underlying server. Developers relying on Langflow should update their installations and avoid exposing administrative interfaces directly to the public internet.

Source: The Hacker News

What to do this week

  • Patch self-hosted GitLab instances and Ivanti Connect Secure gateways against actively exploited flaws.
  • Review and limit third-party JavaScript dependencies on production websites to avoid supply chain script injections.
  • Update Langflow and ensure all internal AI agent workflows are protected behind strict authentication controls.

Stories this brief is based on