LIVE · cybersecurity feed
Live wire
AI found more critical bugs in another widely used technology running 225M websitesCVE-2026-129448 · Apache Roller RCE Flaw CVE-2026-129448 Under Active AttackCognizant and Qualys Redefine AI-Era Managed Security ServiceJava4Shell Vulnerability Allows Remote Attacks on Major Java AppsCVE-2026-114938 · Java4Shell (CVE-2026-114938): Frequently asked questions about OpenJDK remote code execution vulnerabilityCVE-2026-114938 · CVE-2026-114938: Critical vulnerability in OpenJDK JAVACritical Java4Shell flaw allows RCE attacks on major Java applicationsCVE-2026-114938 · How to Find, Mitigate, and Patch the Critical Java4Shell Flaw (CVE-2026-114938) NowCVE-2026-28160 · e4jvikwp VikRentCar Vulnerability Listed As Exploited In VulnCheck KEVCVE-2025-5569 · Ideacms Improper Neutralization Flaw Added To VulnCheck KEV After 514 Days
patch

40% of organizations struggle to retrace AI agents’ actions

A significant portion of modern enterprises is struggling to maintain oversight of autonomous systems, with 40% of organizations reporting an inability to retrace the actions taken by AI agents within their environments. This visibility gap is particularly pronounced in the financial sector, where 46% of firms report difficulty reconstructing agent activity. The challenge is largely attributed…

ZeroDay News ·

Source: Help Net Security

Photo: Namaste jinx (CC BY-SA 4.0) via Wikimedia Commons

A significant portion of modern enterprises is struggling to maintain oversight of autonomous systems, with 40% of organizations reporting an inability to retrace the actions taken by AI agents within their environments. This visibility gap is particularly pronounced in the financial sector, where 46% of firms report difficulty reconstructing agent activity. The challenge is largely attributed to the common practice of logging agent actions under the identities or personal access tokens of human developers, which obscures the distinction between human-led tasks and autonomous machine operations.

According to data from the Palo Alto Networks State of Cloud Security Report 2026, nearly all organizations—99%—now utilize AI agents to perform autonomous or semiautonomous tasks within their development pipelines. These agents have moved beyond simple automation, frequently taking on roles that include designing architectures and deploying applications directly to production. Approximately 75% of organizations have granted these agents permissions to modify code, alter CI/CD configurations, or change production environments.

The rapid pace of AI-driven development has led to an 82% increase in the volume of updates pushed to cloud infrastructure. This velocity often outstrips existing security controls, as many organizations prioritize release speed over rigorous predeployment inspection. Currently, only 37% of organizations enforce manual security reviews, and just 33% possess the capability to trace code from the initial commit to the cloud runtime across the majority of their production workloads. Consequently, roughly 19% of high- and critical-severity application security issues have successfully reached production environments within the last 90 days.

Business pressures, including revenue targets and strict customer deadlines, act as primary inhibitors to security remediation. Roughly 39% of organizations cite release pressure as a major barrier to fixing security issues before deployment, while 35% express concern that security measures could impede development speed. Furthermore, 30% of organizations lack clearly defined risk acceptance criteria, leading to situations where known vulnerabilities are deployed without a clear understanding of ownership or remediation timelines.

The integration of AI agents has also introduced new risks regarding ungoverned workflows. In the past year, 43% of organizations discovered AI agents or automated processes operating without formal security approval. While 97% of organizations express at least moderate confidence in their ability to secure their AI supply chain, this perception is contrasted by limited technical capabilities: only 36% can monitor AI behavior in production, and only 30% are equipped to detect prompt injection attacks. Additionally, 70% of organizations remain unable to map how sensitive data flows through their AI systems.

Identity management remains a critical point of failure. Because 53% of engineers log agent activity using their own credentials, revoking a compromised token becomes a complex task that risks disrupting legitimate services. On average, it takes organizations 2.5 days to rotate or revoke compromised API keys or pipeline secrets. This delay is compounded by a lack of unified telemetry; only 19% of organizations have integrated their workflows across code, cloud, identity, and runtime environments.

Without this unified context, security teams are often forced to manually correlate application vulnerabilities with runtime threats. This inefficiency leads many teams to spend significant time addressing non-exploitable flaws, as they lack the visibility to determine which vulnerabilities are actually exposed to active threats. As frontier AI models become more capable of identifying zero-day vulnerabilities and chaining logic flaws, 65% of organizations now view these advanced AI capabilities as a significant or severe threat to global critical infrastructure and financial systems in the near term.

patchaicloud
ShareXLinkedInWhatsAppFacebook

More News

view all →
vulnerabilitycritical

AI found more critical bugs in another widely used technology running 225M websites

F5 has issued an emergency security update for the open-source NGINX web server to address 10 newly discovered vulnerabilities. These flaws affect a significant portion of the internet, as NGINX currently powers approximately 225 million websites. The vendor has urged all administrators to upgrade their systems to version 1.31.9 immediately to mitigate the risks.

vulnerability

Consensus mounts that tech outages were fueled by patching of AI-discovered vulnerabilities

A series of significant technical outages affecting aviation, banking, and communications infrastructure over the past week has prompted a federal investigation into the stability of critical systems. While initial concerns focused on the possibility of coordinated cyberattacks, officials from the Trump administration, including Secretary of Homeland Security Markwayne Mullin, Director of…

CVE-2026-129448critical

Apache Roller RCE Flaw CVE-2026-129448 Under Active Attack

The Apache Software Foundation has issued an urgent security advisory regarding a critical remote code execution (RCE) vulnerability in Apache Roller, an open-source Java-based blogging platform. The flaw, assigned the identifier CVE-2026-129448 and carrying a maximum CVSS severity score of 9.8, is currently being exploited in the wild.

ai

Coinbase blocks $4.7 million in transfers to Elon Musk deepfake scam

Coinbase successfully intercepted $4.7 million in cryptocurrency transfers associated with a sophisticated deepfake scam targeting users during a high-profile Tesla event. The incident occurred during a demonstration of Tesla's flying car technology, which was broadcast via a fraudulent YouTube livestream. The stream, which drew an audience of over 500,000 viewers, utilized AI-generated video…

ai

Strike Graph Launches Integration AI to Automate Continuous Compliance Evidence Using Any Existing Tool

Strike Graph has announced the launch of a new agentic AI solution designed to automate the collection of compliance evidence across an organization’s existing software stack. By leveraging direct API integrations, the platform aims to streamline the evidence-gathering process, mapping data points to more than 30 different compliance frameworks simultaneously to eliminate redundant manual…

ai

Judge signs off on sale of Spirit Airlines’ data to Google

A federal bankruptcy judge in New York has officially approved the sale of Spirit Airlines’ corporate data to Google, a transaction valued at $10 million. The deal, which follows the airline’s cessation of operations in May, involves the transfer of a massive dataset to the tech giant for the stated purpose of training artificial intelligence models. The repository includes approximately 100…