AI coding agents keep repeating decade-old security mistakes
Coding agents are now writing production features on real development teams, and a new report from DryRun Security shows that those agents introduce security vulnerabilities…
Category: HelpnetSecurity
Authorities dismantle SocksEscort proxy network behind millions in fraud
SocksEscort, a residential proxy network used to exploit thousands of compromised home routers worldwide and facilitate large-scale fraud that cost victims millions of dollars, has…
ShinyHunters claims new campaign targeting Salesforce Experience Cloud sites
Salesforce customers have, once again, been targeted by the ShinyHunters group – or, at least, it’s what the group claims. Attackers modified and abused benign…
Researchers uncover AI-powered vishing platform
A vishing-as-a-service platform that helps scammers carry out so-called “press 1” scams is misusing text-to-speech (TTS) capabilities provided by AI voice technology company ElevenLabs, Mirage…
Virtana enables full-stack root cause analysis beyond legacy APM
Virtana has launched an Application Observability offering that traces performance issues from application code through infrastructure, networks, storage, and AI workloads to deliver evidence-based root…
Meta turns to AI to sniff out scams on Facebook, Messenger and WhatsApp
Meta’s new tools on Facebook, Messenger, and WhatsApp protect users from scams. They use advanced AI systems to analyze text, images, and surrounding context and…
Anthropic forms institute to study long-term AI risks facing society
Anthropic has established the Anthropic Institute, a research unit focused on studying the societal effects of AI and informing policy responses to risks from more…
Microsoft patches 80+ vulnerabilities, six flagged as “more likely” to be exploited
On March 2026 Patch Tuesday, Microsoft addressed 80+ vulnerabilities affecting its software and cloud services. Of these, two were publicly disclosed, but not actively exploited.…
YouTube draws a line on deepfakes involving politicians and journalists
With deepfakes becoming more common, YouTube has expanded access to its AI-driven likeness detection system to a pilot group of government officials, journalists and political…
Armis improves vulnerability accuracy and speed with unified real-time visibility
Armis has announced Armis Centrix for Vulnerability Management Detection and Response. The solution enables security teams to identify and validate vulnerabilities across all organizational assets…
Cloud-audit: Fast, open-source AWS security scanner
Running AWS security audits without a dedicated security team typically means choosing between enterprise platforms with per-check billing and generic open-source scanners that produce findings…
Software vulnerabilities push credential abuse aside in cloud intrusions
Cloud intrusions are unfolding on shorter timelines, with attackers leaning more on unpatched software and compromised identities. H2 2025 distribution of initial access vectors exploited…