From The Hacker News · 40 stories
Security Teams Face Challenges Managing 'Wild Code' Created by Employees
The proliferation of employee-built AI-assisted workflows and applications, termed 'wild code,' is creating new security risks for organizations. These tools often operate with excessive permissions, expose sensitive data, or lack clear ownership, posing challenges beyond traditional software security concerns.
Salesforce Agentforce Vulnerability Exposes Internal Agent Instructions to External Users
Salesforce's Agentforce, when combined with Email-to-Case and Web-to-Case features, creates a path for unauthenticated external users to access and influence internal agent instructions. This vulnerability arises from the interaction between existing platform features and new agent capabilities, rather than a misconfiguration of the agent itself. The issue highlights a gap in governance and security reviews for agents integrated into established platforms.
SANS Introduces Frameworks for AI in Digital Forensics and Incident Response
SANS has developed two new AI frameworks for digital forensics and incident response to address the growing gap between AI tool confidence and human verification capabilities. These frameworks emphasize human oversight as the final authority, despite AI's ability to accelerate investigative work. The initiative aims to prevent unverified AI findings from impacting legal and forensic outcomes.
CISO Report: Cyber Risk Shifts to Workflow, AI Governance, and Human Factors
The 2026 Voice of the CISO report indicates a shift in cyber risk, moving closer to daily workflows, AI governance, and human risk factors. This change requires security leaders to focus on protecting sensitive data within people, cloud platforms, collaboration tools, SaaS applications, and AI-enabled workflows, rather than solely on external threats.
Credential Layer Expansion Outpaces Security Visibility Amidst Accelerating Software Production
The rapid acceleration of software production, evidenced by a significant increase in GitHub commits, is causing the credential layer to expand faster than security teams can monitor it. This growth leads to more exposed credentials, with GitGuardian detecting 28.65 million new hardcoded secrets in public GitHub commits in 2025, highlighting a critical need for enhanced visibility and security measures.
Financial Services Face Increased Software Supply Chain Vulnerability Risks
Financial services companies are facing new pressures to modernize their software supply chains due to advanced AI models that can rapidly exploit dormant vulnerabilities. Historically, these firms accepted vulnerability backlogs for stability, but exploitation has now surpassed phishing as the primary initial access vector for breaches in the sector. This shift necessitates a re-evaluation of long-standing risk management strategies.
Browser-Based Attack Techniques Evolve with Phishing and Malicious Copy-Paste
Browser-based attacks are increasing, with phishing and malicious copy-paste (ClickFix) identified as leading threats. These techniques bypass traditional security measures by intercepting live sessions, using diverse delivery methods, and exploiting user actions to install malware.
DORA's Second Year Focuses on Practical ICT Incident Response and Risk Supervision
The Digital Operational Resilience Act (DORA) is now in its second year of enforcement within the EU, shifting regulatory focus from initial setup to demonstrating effective implementation of ICT incident analysis and risk supervision. Financial entities must ensure their Security Operations Centers (SOCs) have sufficient visibility to detect and investigate intrusions across critical systems, particularly in areas like legacy infrastructure and unmanaged devices. This matters because DORA mandates continuous monitoring beyond basic asset inventories to recognize deviations from normal operational patterns and minimize ICT risk.
Abandoned CDN Domain Re-registered, Posing Supply Chain Risk to Thousands of Sites
An expired CDN domain was re-registered in July 2025, and thousands of websites still hard-code references to its hostnames, allowing the new owner to control content delivered to those sites. This highlights a supply chain vulnerability where third-party scripts can be compromised without a server breach, as seen with the polyfill.io domain in June 2024.
Security Testing Needs to Shift from Individual Techniques to Attack Chains
Current security testing methods, which focus on individual techniques, fail to address how real attackers chain multiple steps together to breach systems. This gap between testing techniques and testing full attack chains leaves organizations vulnerable, despite validated individual controls. The industry needs to adapt its testing strategies to simulate multi-stage attacks to effectively counter modern threats.
F5 BIG-IP APM Malware Hides PHP Web Shell in Memory, Evading Disk Scans
Sophos reported that malware targeting F5 BIG-IP Access Policy Manager appliances injects a PHP web shell directly into memory, bypassing traditional disk-based file scans. This technique allows attackers to maintain persistence and execute commands without leaving traces on the file system, making detection more challenging for organizations using these F5 products.
Cloud Security Index 2026 Reveals Divergent Misconfiguration Risks Across AWS, Azure, GCP
Intruder's 2026 Cloud Security Index analyzed misconfiguration data from 3,000 organizations across AWS, Azure, and Google Cloud, finding that risk profiles vary significantly between providers. This matters because security checklists need to be tailored to specific cloud platforms rather than using a one-size-fits-all approach.
AI-powered agents are transforming phishing attacks, increasing sophistication and scale
Phishing attacks are evolving into "Phishing 3.0," where AI agents automate reconnaissance, lure generation, and multi-channel delivery, making attacks more sophisticated and scalable. This shift means traditional email defenses are becoming obsolete as the threat moves beyond simple malicious content to AI-driven social engineering across various communication platforms.
Picus Labs Report: Enterprise Perimeter Defenses Improve, Internal Security Weak Against Covert Attacks
Picus Labs' Blue Report 2026 indicates that enterprise perimeter defenses have improved, with prevention effectiveness reaching 69% in the first half of 2026. However, internal network security remains weak, particularly against quiet attack techniques like reconnaissance and credential theft, which are stopped less than 22% of the time. This disparity means that while external threats are better mitigated, attackers who bypass the perimeter face little resistance once inside the network.
TeamPCP Linked to Redis Attacks Since 2020 and Supply Chain Campaigns
New analysis by Oligo Security researchers links the threat actor TeamPCP to Redis server compromises dating back to 2020, predating their known supply chain attacks. This connection is based on overlapping infrastructure, malware deployment, and operational techniques, indicating a long-standing presence in cybercrime before targeting software supply chains.
CTM360 Research Details Real-Time Account Hijacking in Insurance Phishing Campaigns
CTM360 research indicates a shift in insurance phishing tactics from credential harvesting to real-time account hijacking, where attackers authenticate immediately as victims log in. This evolution allows attackers to gain immediate access to sensitive personal and financial data, posing a greater risk than traditional phishing methods.
Synthetic Identity Fraud Emerges as Threat to Machine Identities
Synthetic identity fraud, traditionally understood as creating fake human identities, is now identified as a significant, under-discussed threat to Non-Human Identities (NHIs). Attackers can fabricate new machine identities by blending real and fake attributes, making them difficult to distinguish from legitimate ones. This method bypasses traditional security measures focused on detecting hijacked accounts, posing a challenge for organizations managing a growing number of NHIs.
Recent Clearinghouse Announcements Lack Significance, Says Industry Expert
Numerous tech companies have launched clearinghouses for vulnerability data, including Athena. The article argues that while the influx of clearinghouses may seem significant, most won't provide actionable solutions to security issues in the software supply chain.
AI Integration Changes Software Supply Chain Security Approaches
The integration of AI in the software supply chain is altering security paradigms by shifting risk away from solely code to involve AI-generated dependencies and tools. This necessitates enhanced governance and validation processes to manage the evolving threat landscape associated with AI components.
Challenges of Identity Lifecycle Management for AI Agents
Identity lifecycle management systems, designed for human employees, struggle to accommodate AI agents. This gap presents governance issues as enterprises increasingly integrate autonomous agents, necessitating updates to existing frameworks.
Richard Bejtlich Advocates for NDR in Modern Security Operations
Richard Bejtlich highlights the growing need for Network Detection and Response (NDR) in cybersecurity. His guide emphasizes moving beyond traditional alerts to prioritize actionable evidence in detecting and mitigating threats.
OpenAI Employee Pleads Not Guilty to Planting Listening Device
Abdulrahman Alfozan, an OpenAI employee and venture partner at Raed Ventures, pleaded not guilty to charges of planting a listening device in an OpenAI executive conference room. He faces a maximum of five years in prison if convicted, amidst growing concerns over AI technology theft.
Meta Patents AI to Monitor Emotions and Fitness via Voice Recordings
Meta has patented a system that uses AI to monitor users' emotions by analyzing their voice and contextual data. The technology tracks emotional states and links them to physical location and activity for potential use in personalized fitness coaching. This patent showcases Meta's continued interest in integrating user data with AI for customized experiences.
Bluesky Phishing Attack Compromises Over 4,500 Accounts; Singapore Police Warn of Android Malware Scam
Bluesky reported a phishing attack that compromised over 4,500 user accounts through fake government payment promotions. Separately, the Singapore Police Force warned of an Android malware scam, resulting in S$989,000 in losses from 62 reported cases since April 2026.
Chainguard Reaches 1 Billion Container Build Manifests, Doubles Output in Six Months
Chainguard announced it has doubled its container build manifest output to over 1 billion in the last six months, now supporting more than 3,000 unique container images and 675,000 image versions. This milestone reflects the company's continuous rebuilding process for maintaining up-to-date and secure container images through its Chainguard OS and Factory infrastructure.
Integrating Autonomous AI and Analyst Copilots in Security Operations Centers
A Fortune 50 CISO discussed AI agents in security operations centers (SOC), revealing limitations in current designs that may overlook many alerts requiring human judgment. The insights of psychologist Daniel Kahneman emphasize the need for a balanced approach between automatic and deliberate human cognition in AI architecture for effective security.
Rich Network Data Improves Security Incident Investigation Beyond Basic Alerts
Security investigations often hit dead ends when relying solely on basic alert data. Rich network data, which includes full transaction details and context, allows analysts to trace attacker activity, understand the scope of an incident, and intervene effectively. This approach is gaining importance as AI defense systems also rely on high-quality data.
Safe Remediation Method Addresses Delays in Applying Urgent Security Fixes
A method called "Safe Remediation" is proposed to address delays in applying urgent security fixes by providing evidence of changes before implementation. This approach aims to build confidence among operations teams that security actions will not disrupt services. It involves four stages: Observe, Prove, Act, and Confirm, using existing data and testing to validate proposed changes.
Attackers Operate in Gaps Between Security Alerts, Exploiting SOC Operational Limits
Attackers exploit the time between security alerts and blend into normal operations to evade detection, making it difficult for SOC teams to trace the full scope of an intrusion. Modern attacks use AI for speed but also employ stealth tactics like slow reconnaissance and using standard admin tools to remain undetected for extended periods.
Deepfake Readiness Drills Address Rising Voice and Video Cloning Attacks
Organizations are increasingly facing deepfake attacks involving cloned voices and synthetic video calls, used for payment fraud and unauthorized access. Implementing deepfake readiness drills allows security teams to identify vulnerabilities in verification processes before real incidents occur. This approach shifts defense from relying on auditory detection to strengthening procedural safeguards against sophisticated impersonations.
CISOs struggle to answer board questions on security posture and financial exposure
Many CISOs cannot confidently answer board questions regarding overall organizational security, financial exposure, and security posture improvement due to fragmented data across multiple security tools. Boards require reporting focused on exposure, trend, and financial impact rather than activity metrics to assess risk effectively.
Identity Visibility is Critical for Modern Identity Security
Identity visibility, which involves seeing every identity, its access, and how that access is used, is foundational for modern identity security. This is crucial because stolen credentials are a frequent initial access vector in data breaches, and cloud environments complicate identity management by creating "identity dark matter."
Autonomous Penetration Testing Prioritizes Vulnerabilities Based on Attack Paths
Security teams are shifting focus from isolated vulnerability severity to understanding actual attack paths to compromise. Autonomous penetration testing provides continuous validation by simulating attacker actions to reveal exploitable weaknesses in complex environments. This approach helps prioritize remediation based on real-world risk rather than theoretical severity scores.
Identity Fabric Architecture Addresses Fragmented Identity Systems in Hybrid Cloud Environments
An Identity Fabric is an architectural approach that unifies fragmented identity systems across applications, APIs, and infrastructure, providing runtime visibility into identity behavior. This approach addresses the increasing complexity of identity management in hybrid and multi-cloud environments, where traditional IAM tools struggle to verify access implementation. It matters because it closes the gap between intended access policies and actual access execution, reducing risks associated with unobserved identity activity.
Frontier AI Models Transform Vulnerability Management Landscape
Frontier AI models, such as Anthropic's Mythos, are changing vulnerability management by identifying zero-day flaws and chaining complex exploits. This development necessitates a systemic revolution in how organizations approach vulnerability and patch management. The traditional methods of prioritizing vulnerabilities using CVSS, EPSS, and CISA's KEV list are becoming insufficient as AI rapidly creates new exploits.
Check Point Introduces AI Network Firewall for AI Security
Check Point has launched the AI Network Firewall, integrating it into its AI Defense Plane to secure AI activity across enterprise networks. This new firewall aims to address the visibility gap created by AI's impact on network dynamics, which traditional firewalls were not designed to handle.
Webinar on Defending Against Rapid AI-Powered Attacks Announced
A free webinar hosted by Zscaler will focus on defending against AI-driven attacks, which have become significantly faster. The session will provide strategies to adapt security measures in response to these advanced threats, emphasizing Zero Trust principles.
Open Source Faces New Challenges from AI and Industrialized Malware
Open source software is transitioning from an unregulated, community-driven model to one facing significant external pressures. This shift is driven by increased security threats from AI-generated zero-days and widespread malware, alongside new regulations and executive orders. The change will impact how enterprises consume open source, though the core definition of Open Source (OSI-stewarded) will remain unchanged.
Guide to IAM Compliance Requirements and Best Practices
This guide explains Identity and Access Management (IAM) compliance, focusing on the distinction between policy intent and actual enforcement of access controls. It details why organizations need to move beyond periodic access reviews to continuous, evidence-backed verification for auditors. The guide highlights common evidence gaps and the importance of verifying implementation over just design.
Wazuh Integrates AI for Enhanced Security Operations Center Workflows
Wazuh has integrated AI capabilities, including the Wazuh AI Analyst, into its security platform to assist Security Operations Centers (SOCs) in managing high alert volumes and complex investigations. This integration aims to augment human analysts by providing contextual explanations, summarizing findings, and recommending remediation actions, thereby improving threat detection and incident response.