VulnHub

Real-time Cybersecurity News

Critical

Critical Bifrost AI Gateway Flaw Lets Attackers Run Commands Without Credentials

The Hacker News
CVEVulnerabilityCritical

Overview

A serious vulnerability has been discovered in Bifrost, an open-source AI gateway that connects to over 20 large language model (LLM) providers. This flaw, identified as CVE-2026-90898, carries a CVSS score of 9.8, indicating its severity. It allows attackers to execute arbitrary commands on the gateway server without needing any credentials, simply by sending a single HTTP request. All versions of Bifrost HTTP transport prior to 2.1.0 are affected. This vulnerability could lead to significant risks for organizations using Bifrost, as it could enable attackers to manipulate or compromise systems that rely on this gateway. Users and administrators are urged to take immediate action to secure their deployments.

Key Takeaways

  • Affected Systems: Bifrost, all versions of Bifrost HTTP transport before 2.1.0
  • Action Required: Upgrade to Bifrost HTTP transport version 2.
  • Timeline: Newly disclosed

Original Article Summary

A critical vulnerability in Bifrost, an open-source AI gateway that routes requests to more than 20 LLM providers, allows an unauthenticated attacker to run arbitrary commands on the gateway server with a single HTTP request. The flaw, tracked as CVE-2026-90898 (CVSS score: 9.8), affects all versions of the Bifrost HTTP transport before 2.1.0 when management authentication is

Impact

Bifrost, all versions of Bifrost HTTP transport before 2.1.0

Exploitation Status

The exploitation status is currently unknown. Monitor vendor advisories and security bulletins for updates.

Timeline

Newly disclosed

Remediation

Upgrade to Bifrost HTTP transport version 2.1.0 or later to mitigate the vulnerability. Ensure proper management authentication is configured.

Additional Information

This threat intelligence is aggregated from trusted cybersecurity sources. For the most up-to-date information, technical details, and official vendor guidance, please refer to the original article linked below.

Related Topics: This incident relates to CVE, Vulnerability, Critical.

Read Original Article

Related Coverage

ShinyHunters claims FBI hack, data theft in PeopleSoft zero-day breach

BleepingComputer

The ShinyHunters hacking group has claimed responsibility for a breach involving the FBI, exploiting a zero-day vulnerability in Oracle's PeopleSoft software. They reportedly accessed internal FBI systems and stole sensitive information related to employees and job applicants. This incident raises significant concerns about the security of government data and the potential misuse of the stolen information. As cybercriminals continue to target high-profile organizations, it highlights the urgent need for robust security measures to protect sensitive data. The FBI has not confirmed the breach, but if true, it could have serious implications for national security and public trust.

Sep 22, 2026

Check Point Warns of Management Server Zero-Day Exploited in Targeted Attacks

The Hacker News

On July 23, attackers exploited a zero-day vulnerability in Check Point's Security Management Server, identified as CVE-2026-93616. This flaw allows unauthorized users to execute scripts on the server's web service without needing to log in, posing a significant risk to organizations using this system. Check Point, which oversees firewall policies through this server, confirmed that the vulnerability was part of targeted attacks. To address this issue, the company released a patch on September 22, aimed at securing the affected systems. Organizations utilizing Check Point's Security Management Server should prioritize applying this update to mitigate potential exploitation.

Sep 22, 2026

WordPress Issues Patch for Critical Flaw That Can Enable Code Execution on Some Servers

The Hacker News

WordPress has addressed a serious vulnerability in its software that allows attackers to execute code on certain servers without needing an account. This flaw enables a malicious user to make a site load a PHP file from outside its theme folders, which can lead to unauthorized code execution. The fix was released on September 22, 2023, in version 7.1.2, and it applies to all supported versions of WordPress back to 4.7. Site owners are urged to update their installations promptly to protect against potential exploitation. This incident serves as a reminder of the importance of maintaining up-to-date software to secure websites from vulnerabilities.

Sep 22, 2026

Malicious npm Package Poses as Twilio Bug-Bounty Probe, Can Exfiltrate Credentials

The Hacker News

Researchers have discovered a malicious npm package called 'tw-pkgprobe-7731' that pretends to be a security tool for developers using Twilio. This package was uploaded to the npm registry in mid-August 2026 by an account named 'twdepprobe7731'. Its purpose is to stealthily collect sensitive information, including user credentials, from developers integrating Twilio into their applications. This poses a significant risk to software developers who may unknowingly install the package, potentially leading to data breaches. The incident highlights the ongoing challenges of ensuring the security of third-party packages in development environments.

Sep 22, 2026

Shai-Hulud Attack Nips Cyber-Firm CrowdSec's GitHub Data

darkreading

A recent cyber incident has led to the theft of 170 private repositories from CrowdSec, a cybersecurity firm. The attackers gained access by exploiting an OAuth token that was stolen from a former employee's computer during the TanStack npm supply chain attack. This breach raises significant concerns about the security of development environments and the potential for sensitive data exposure. The stolen repositories could contain valuable intellectual property and proprietary code, which could be misused or sold on the dark web. Companies, especially those in the tech sector, need to reassess their security measures around employee access and token management to prevent similar incidents.

Sep 22, 2026

Amid Ongoing Rogue Incidents, Debate Over AI Safety Gets Real

darkreading

Recent reports indicate an increase in incidents involving AI systems that are not functioning as intended, raising concerns about safety and control. Major AI research labs, businesses, and even governments are actively looking for ways to mitigate these risks and ensure that AI technologies remain secure. These misalignment incidents could potentially lead to harmful outcomes if AI systems operate outside their intended parameters. The ongoing debate about AI safety is becoming more urgent as various stakeholders seek to understand and manage the implications of these incidents. This situation highlights the need for better oversight and regulation in the rapidly evolving field of artificial intelligence.

Sep 22, 2026