A week in security (July 27 – August 2)

Last week on Malwarebytes Labs: Fake Fortnite rewards are stealing players’ accounts Fake Flash Player installs AtlasRAT Malwarebytes for Windows, now available on the Microsoft Store Hims & Hers sued over alleged health data privacy failures Hidden prompt turns Microsoft Copilot into an AI worm...

The OpenAI and Anthropic AI Hacking Sprees Are a Messy New Legal Frontier

Both major AI labs’ models broke containment, escaped onto the internet, and hacked other companies. If a human had done that, the law would likely be against them. But a bot?...

openSUSE 16 Security Update : perl-Mojolicious (openSUSE-SU-2026:21496-1)

The remote openSUSE 16 host has a package installed that is affected by a vulnerability as referenced in the openSUSE- SU-2026:21496-1 advisory. Changes in perl-Mojolicious: - updated to 9.480.0 9.48 - Fixed a security issue where CSRF tokens were vulnerable to BREACH attacks. Tokens are now mask...

HollowFrame Loader Deploys Matryoshka Backdoor in Spear-Phishing Attack on Law Firm

Cybersecurity researchers have shed light on a previously undocumented Go-based loader framework called HollowFrame and a Rust-based malware family tracked as Matryoshka. According to Blackpoint Cyber, the intrusion sequence begins with a spear-phishing message containing a link to an encrypted...

EUVD-2026-51480

Keycloak provides a way to let users log in using Microsoft accounts while restricting access to a specific organization tenant. A flaw was discovered where this restriction is ignored when using the token exchange feature. This means an attacker with a valid Microsoft token from a completely...

Anthropic Says Claude Mistook the Open Internet for a CTF and Breached Three Organizations

Anthropic on Thursday became the latest artificial intelligence AI company to reveal that three of its models, including Claude Opus 4.7, Mythos 5, and an unnamed research model, had breached three unnamed organizations during cybersecurity testing without its knowledge. The AI firm said the...

CVE-2026-58039

A flaw in Node.js Permission Model enforcement allows process.report writes and overwrites files outside --allow-fs-write paths. This can lead to confidentiality impact or bypass of the intended security boundary under affected configurations. This vulnerability affects Node.js 22.x, 24.x, and 26...

Security update for perl-Mojolicious (moderate)

openSUSE security update: security update for perl-mojolicious ------------------------------------------------------------- Announcement ID: openSUSE-SU-2026:21496-1 Rating: moderate References: bsc1271431 Cross-References: CVE-2026-15747 CVSS scores: CVE-2026-15747 SUSE : 5.9...

Malicious code in 1apijs (npm)

--- -= Per source details. Do not edit below this line.=- Source: ghsa-malware e2700f92a449d39fb0881259de81734a6c2f7933dad7e567fd4284e58b6fe062 Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be...

MAL-2026-11267 Malicious code in compose-logger-stand (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector a34de797bc2020fd52b9b36fda8513a111610612c3a50e75d8ea8a79440be4f0 [email protected] presents itself as a logger but its shipped code implements a credential-harvest and exfiltration pipeline...

CVE-2026-62246 Kamaji: TenantControlPlane namespace/name collision binds two tenants to the same SQL datastore schema + DB user, breaking per-tenant isolation

Kamaji is the Hosted Control Plane Manager for Kubernetes. Prior to 26.7.4-edge, Kamaji derives a TenantControlPlane datastore schema, database user, and etcd key prefix from a lossy namespace-and-name normalization in GetDefaultDatastoreSchema and GetDefaultDatastoreUsername, allowing distinct...

CVE-2026-62246 Kamaji: TenantControlPlane namespace/name collision binds two tenants to the same SQL datastore schema + DB user, breaking per-tenant isolation

Kamaji is the Hosted Control Plane Manager for Kubernetes. Prior to 26.7.4-edge, Kamaji derives a TenantControlPlane datastore schema, database user, and etcd key prefix from a lossy namespace-and-name normalization in GetDefaultDatastoreSchema and GetDefaultDatastoreUsername, allowing distinct...

CVE-2026-48499 Activepieces: Cross-tenant data exposure and code injection via the Code piece sandbox cache

Activepieces is an open source AI workflow automation platform. Prior to 0.84.0, an unsanitized path segment in the Code piece sandbox can let an authenticated flow author reach read-write cached flow and code files belonging to other tenants on the same worker, exposing embedded data and allowin...

CVE-2026-10700

IBM Langflow OSS 1.0.0 through 1.8.4 contains multiple broken access control vulnerabilities in its file handling API that allow unauthorized access to user files.The /api/v1/files/images/flowid/filename endpoint does not enforce authentication or authorization checks, allowing unauthenticated...

CVE-2026-44101

Due to missing authentication the CHARX OCPP Agent service allows an unauthenticated remote attacker to reconfigure the backend connection. This can lead to Denial-of-Service and confidential data being disclosed to the attacker...

CVE-2026-64460

A flaw was found in the Linux kernel's PCI/IOV subsystem. When a PCI device fails to respond during a power state transition, the kernel's Virtual Function VF Resizable Base Address Register BAR restoration process can attempt to access memory outside of its allocated bounds. This out-of-bounds...

LeakNet Claims 11TB of Data Stolen in NYC Health + Hospitals Breach

LeakNet claims it stole 11TB of NYC Health + Hospitals data containing sensitive medical, financial and biometric records linked to more than 12 million people...

Measuring the Tendency of AI Agents to Go Rogue

This essay was written with Barath Raghavan, and originally appeared inThe Guardian. In July, Hugging Face, a company that hosts much of the world's AI software and open-source AI models, was hacked. A malicious dataset had been used to run code on one of its servers. Whoever was behind it captur...

OPENSUSE-SU-2026:21496-1 Security update for perl-Mojolicious

This update for perl-Mojolicious fixes the following issues: Changes in perl-Mojolicious: - updated to 9.480.0 9.48 - Fixed a security issue where CSRF tokens were vulnerable to BREACH attacks. Tokens are now masked with a fresh random value on every request, instead of being reused for the whole...

OpenAI Agent Used Exposed Credentials Across Four Services During Hugging Face Breach

OpenAI on Tuesday revealed the rogue artificial intelligence AI agent that escaped its sealed evaluation environment and broke into Hugging Face's production environment also hacked multiple third-party accounts and services as part of the attack. The latest disclosure shows that the security...