News

Researchers at web security company Radware recently discovered what they described as a service-side data theft attack method involving ChatGPT.  The attack, dubbed ShadowLeak, targeted ChatGPT’s Deep Research capability, which is designed to conduct multi-step research for complex tasks. OpenAI neutralized ShadowLeak after it was notified by Radware. The ShadowLeak attack did not require any user interaction. The attacker simply needed to send a specially crafted email that when processed by the Deep Research agent…

Read More

SecurityWeek’s Attack Surface Management Virtual Summit is now LIVE and runs today from 11AM – 4PM ET. Join the online event where cybersecurity leaders and practitioners will dive into the strategies, tools, and innovations shaping the future of ASM. As digital assets and cloud services continue to expand, defenders are shifting tactics to continuously discover, inventory, classify, prioritize, and monitor their attack surfaces. This summit brings together experts to share real-world lessons, emerging trends, and practical…

Read More

CrowdStrike on Tuesday said that it would acquire Pangea, a company specializing in AI security, to expand its Falcon platform with new protections designed for enterprise AI systems. The acquisition, announced at CrowdStrike’s Fal.Con 2025 event, is intended to address security challenges specific to the use of AI models, agents, and applications in the workplace. CrowdStrike plans to integrate Pangea’s capabilities to help organizations monitor, control, and secure AI interactions across their infrastructure. The announcement…

Read More

SecurityWeek’s cybersecurity news roundup provides a concise compilation of noteworthy stories that might have slipped under the radar. We provide a valuable summary of stories that may not warrant an entire article, but are nonetheless important for a comprehensive understanding of the cybersecurity landscape. Each week, we curate and present a collection of noteworthy developments, ranging from the latest vulnerability discoveries and emerging attack techniques to significant policy changes and industry reports.  Here are this…

Read More

Threat actors are exploiting a critical-severity vulnerability in DELMIA Apriso factory software, the US cybersecurity agency CISA warns. Developed by French company Dassault Systèmes, DELMIA Apriso is a manufacturing operations management (MOM) and manufacturing execution system (MES) software designed for managing every detail of the manufacturing process. The software is used in North America, Europe, and Asia, including in the aerospace and defense, automotive, high-tech, and industrial equipment industries.  Tracked as CVE-2025-5086 (CVSS score of…

Read More

Apple in early September sent a fresh wave of threat notifications to French users it believes might have been targeted by commercial spyware. This is at least the fourth time the Cupertino-based tech giant has notified users in France of potential mercenary spyware attacks, according to an alert from the French national Computer Emergency Response Team (CERT-FR). “This alert records all waves of notifications sent by Apple and known to CERT-FR since March 5, 2025.…

Read More

Nevada on Wednesday confirmed that the days-long disruption to state systems and services was caused by a ransomware attack. The incident, disclosed on Monday morning as a network security incident, occurred on Sunday, and forced Nevada to close all state offices on Monday and Tuesday. During a press conference on Wednesday, the state’s officials publicly confirmed that a “sophisticated ransomware attack” was the cause of the disruptions. “Upon detection, we immediately activated our established cybersecurity…

Read More

Two serious vulnerabilities were patched recently by Xerox in its FreeFlow Core print orchestration platform.  According to pentesting company Horizon3, whose researchers discovered the flaws, FreeFlow Core is affected by an XXE injection flaw (CVE-2025-8355) and a path traversal issue (CVE-2025-8356). The researchers discovered that the vulnerabilities could allow an unauthenticated, remote attacker to execute arbitrary code on affected FreeFlow Core instances. The potential impact has been demonstrated with an exploit that placed a webshell…

Read More

Two different firms have tested the newly released GPT-5, and both find its security sadly lacking. After Grok-4 fell to a jailbreak in two days, GPT-5 fell in 24 hours to the same researchers. Separately, but almost simultaneously, red teamers from SPLX (formerly known as SplxAI) declare, “GPT-5’s raw model is nearly unusable for enterprise out of the box. Even OpenAI’s internal prompt layer leaves significant gaps, especially in Business Alignment.” NeuralTrust’s jailbreak employed a…

Read More