A clearer view of today.

New reporting from Ukraine’s SSU and the FBI describes a sustained Russian intelligence campaign targeting the messenger accounts of officials, military personnel, politicians, and activists across multiple regions. The focus on accounts and “recovery” access underscores how attackers can maintain footholds even after victims change devices or switch credentials. For defenders, the takeaway is to treat secure messaging accounts as high-value targets and to harden account recovery paths, not just login credentials.

Source: Security Affairs

The US has announced a major bounty tied to Russian threat activity against US government officials, military leaders, and allied personnel. The move signals increased prioritization of attribution-linked disruption for operational messaging abuse—an area that historically blends phishing, social engineering, and account takeover. Organizations should review whether their threat models include messaging platform access paths (including backups and recovery flows) as part of incident response planning.

Source: SecurityWeek

Recent analysis highlights how prompt injection is evolving from “tell the model to do X” into attacks that manipulate multi-agent systems, retrieval-augmented generation (RAG), and routing logic between multiple models. The core risk is that enterprises often grant LLM-powered systems too much trust—allowing injected instructions to trigger actions, leak data, or corrupt workflows. Defenders should implement permission constraints, treat external RAG sources as untrusted, and add controls around tool invocation and provenance validation.

Source: VentureBeat

A new executive-order analysis argues that federal post-quantum cryptography (PQC) timelines are compressing the window for “orderly execution,” forcing CISOs to plan now. The message is less about pilots and more about multi-year transformation—inventorying cryptography, planning migrations, and aligning stakeholders across engineering, risk, and procurement. Organizations that haven’t started should expect near-term pressure to demonstrate readiness rather than progress.

Source: CyberScoop

A public proof-of-concept is now available for a critical libssh2 flaw (CVE-2026-55200) that can cause memory corruption when a client connects to a malicious or compromised SSH server. Because libssh2 is a client-side library, the practical threat is heightened for environments that routinely connect to untrusted or externally controlled SSH endpoints. If you can’t fully inventory where libssh2 is used, prioritize patching and constraining outbound SSH exposure to known-good systems.

Source: The Hacker News

Microsoft says it shut down a long-running malicious extension operation on the Edge Add-ons store that hid payloads inside image and font files, then activated days after installation. The campaign reportedly stole credentials and supported ad fraud, demonstrating that “benign-looking” assets can be weaponized to bypass simpler scanners. Teams should review extension policies, restrict install privileges, and strengthen monitoring for post-install suspicious behavior (especially delayed execution).

Source: The Hacker News

VirusTotal’s YARA-X project shipped updated releases (1.18.0 and 1.19.0) that include multiple improvements and bug fixes—important for teams relying on YARA-X for detection and hunting workflows. While these releases are defensive and tooling-focused, keeping detection infrastructure current helps maintain reliability against evolving malware techniques and reduces false negatives caused by bugs. Security teams should schedule updates for internal analysis environments and CI detection pipelines.

Source: SANS ISC