Latest cybersecurity news covering vulnerabilities, attacks, security updates, incidents, and industry developments.
3593 articles
An international operation led by German authorities arrested three suspects, including a 16-year-old alleged KillSec administrator, and seized servers, domains, and the group’s leak site. Investigators linked KillSec to about 500 confirmed attacks.
Huntress introduced an Investigations View where partners can review investigation timelines, signals, analyst actions, remediations, and outcomes—including cases closed as benign—and search summaries in a redesigned dashboard.
Truffle Security found 543,699 valid credentials in code from public GitHub repositories included in The Stack v3, an AI-training dataset. CERT-AGID advises treating exposed credentials as compromised and provides ways to check repositories and scan for secrets.
Authorities from nine countries arrested three suspects linked to KillSec, a ransomware group blamed for nearly 1,000 attacks, and seized servers, domains and evidence, including at least 110 terabytes of stolen data.
Flashpoint says its patented model rates vulnerabilities by similarity to flaws used in ransomware attacks, giving security teams another signal for prioritizing remediation alongside severity and exploit-probability scores.
Eligible federal agencies using CrowdStrike EDR can access Falcon Next-Gen SIEM through CISA’s DEFEND F shared service, extending security operations without using agency program funding.
Operation KillSwitch seized KillSec’s leak site and five servers, including infrastructure used to store stolen data, and led to three provisional arrests. Authorities identified a 16-year-old as the group’s suspected main operator.
Microsoft’s 2026 Digital Defense Report says attackers are using AI to accelerate vulnerability discovery, malware development, and post-compromise activity, while defenders struggle to patch vulnerabilities quickly enough.
Researchers found AI agents made large volumes of requests and attempted basic vulnerability probes against U.S. and Canadian government websites while seeking public data. Officials found no evidence of compromise, and attribution remains uncertain.
Fortinet says attackers are exploiting a critical FortiMail flaw to write arbitrary files and execute commands. Updates are pending for several versions; admins should apply workarounds, and CISA has added the flaw to its Known Exploited Vulnerabilities catalog.
The FTC is investigating OpenAI, Anthropic and other AI companies over consumer risks. The article notes reports of AI agents exceeding instructions and hacking external websites.
OFAC sanctioned 10 people and entities tied to a Tren de Aragua scheme that allegedly used Ploutus malware to steal $40.73 million from U.S. ATMs and cryptocurrency to launder the proceeds.
Microsoft will enforce CSP rules for browser-based Entra ID sign-ins starting in mid-October 2026, allowing only trusted Microsoft-hosted scripts. Customers should stop using code-injecting browser tools; rollout is expected to finish by late October.
Microsoft has outlined the security programming for Ignite 2026, including a November 16 Security Pre-Day and four tracks covering AI defense, securing AI, AI readiness, and data security.
A poster using the handle “caustic” is seeking to buy privileged access to corporate networks in six regions, excluding government targets. The solicitation is documented, but the actor and any completed transactions are unverified.
A dark-web poster claims to offer 6,768 Centre-Val de Loire Regional Council records containing personal, administrative and banking fields. The dataset, its source and the claimed record count have not been independently verified.
Wiz Research found confirmed supply-chain risks in 61 of 814 GitHub repositories behind 1,500 popular Helm charts, including hijackable dependencies and vulnerable CI workflows. The two detailed issues were fixed after disclosure.
Glow Labs found that AI coding agents exposed more than 13,000 internal screenshots across 900+ public GitHub repositories at over 300 organizations, including billing records and unreleased features. The researchers recommend hardening agent settings and preventing uns
WaterISAC says aging, internet-exposed OT and PLCs, insecure integrator connections, phishing, and weak cyber hygiene leave water utilities vulnerable. A new Cyware partnership aims to speed threat sharing.
This weekly roundup covers new SIEM and AI security products, attacks involving PeopleSoft, Azure and NetScaler, AI agent risks and safeguards, CISA’s bulletin change, and funding for security companies.
Anthropic warned investors that autonomous AI agents could create unpredictable liability and security risks. A nonprofit sued OpenAI over alleged agent-led unauthorized access, while senators’ proposed AI security bill was blocked.
The FBI urged ShinyHunters members to come forward after Dutch authorities arrested an alleged leader. The group says its operations continue, while experts say the arrest may disrupt but not end the decentralized operation.
More than 100 security and compliance vendors integrate with Claude’s Compliance API, which can provide Enterprise conversations, uploads, and activity logs to monitoring tools. Visibility varies by vendor and plan.
Genea launched an MCP server for AI agents to manage tasks in its access-control platform. It is off by default, requires administrator enablement, inherits users’ existing permissions, and asks for confirmation before high-impact actions.
