Latest cybersecurity reporting from selected sources.

AutoJack is a novel exploit chain showing how a single malicious webpage can turn an AI browsing agent into a remote code execution vector on the host machine. By abusing trust in localhost, missing authentication, and unsafe parameter handling, attackers can trigger arbitrary process execution through AutoGen Studio’s MCP WebSocket. The research highlights a broader pattern - when agents can browse untrusted content and access local services, traditional boundaries like localhost are no longer secure. The post AutoJack: How a single page can RCE the host running your AI agent  appeared first on Microsoft Security Blog.

Seeking to Add Urgency, Mandiant Publishes Rainbow Tables for NTLM Key HashesFor nearly 30 years, security experts have warned organizations to ditch the weak NTLM authentication protocol in Windows. But its use persists, even amidst easy and active exploits. Now Google has published rainbow tables for NTLMv1. Will this finally drive holdout organizations to change?

Microsoft on Monday attributed a threat actor it tracks as Storm-1175 to the exploitation of a critical security flaw in Fortra GoAnywhere software to facilitate the deployment of Medusa ransomware

The recent mass-theft of authentication tokens from Salesloft, whose AI chatbot is used by a broad swath of corporate America to convert customer interaction into Salesforce leads, has left many companies racing to invalidate the stolen credentials before hackers can exploit them. Now Google warns the breach goes far beyond access to Salesforce data, noting the hackers responsible also stole valid authentication tokens for hundreds of online services that customers can integrate with Salesloft, including Slack, Google Workspace, Amazon S3, Microsoft Azure, and OpenAI.

Microsoft Issuing Emergency Patches to Combat Authentication-Bypassing AttacksHackers have been exploiting a zero-day vulnerability dubbed "ToolShell" in on-premises installations of Microsoft SharePoint to gain remote access, and steal cryptographic keys and data. As Microsoft rolls out patches, experts warn administrators to also rotate keys, to help eject attackers.

Multiple suspected Russia-linked threat actors are "aggressively" targeting individuals and organizations with ties to Ukraine and human rights with an aim to gain unauthorized access to Microsoft 365 accounts since early March 2025

A proof-of-concept (PoC) attack vector exploits two Azure authentication tokens from within a browser, giving threat actors persistent access to key cloud services, including Microsoft 365 applications.

Proof-of-concept exploit code is now public for a vulnerability in Microsoft's Remote Registry client that could be used to take control of a Windows domain by downgrading the security of the authentication process. [...]

Phemedrone Stealer loots drives for passwords, cookies, login tokens, etc Criminals are exploiting a Windows Defender SmartScreen bypass vulnerability to infect PCs with Phemedrone Stealer, a malware strain that scans machines for sensitive information – passwords, cookies, authentication tokens, you name it – to grab and leak.…

Cybersecurity researchers have discovered a case of "forced authentication" that could be exploited to leak a Windows user's NT LAN Manager (NTLM) tokens by tricking a victim into opening a specially crafted Microsoft Access file

Proof-of-concept exploit code has surfaced on GitHub for a critical authentication bypass vulnerability in Microsoft SharePoint Server, allowing privilege escalation. [...]

The Chinese state-sponsored group known as UNC3886 has been found to exploit a zero-day flaw in VMware ESXi hosts to backdoor Windows and Linux systems

Two of the vulnerabilities — in Azure Functions and Azure Digital Twins — required no account authentication for an attacker to exploit them.

The massive phishing campaign does not exploit a vulnerability in MFA. Instead, it spoofs an Office 365 authentication page to steal credentials.