Cisco warns of critical vulnerabilities in Secure Firewall ASA and FTD software, exposing unpatched systems to denial-of-service and potential remote code execution. Discovered by security researchers, the flaws are actively exploited, urging swift patching to prevent sophisticated threats.

Cisco's Unified Contact Center Express is vulnerable to critical Java RMI flaws, risking remote code execution and privilege escalation, with a CVSS score of 9.8. Cisco urges immediate updates, as no workarounds exist, though no active exploitation has been reported yet.

A sophisticated Iranian-linked cyber campaign, dubbed UNK_SmudgedSerpent, uses social engineering and remote management tools to target academics and policy experts. The operation, active from June to August 2025, features phishing tactics mimicking legitimate platforms to harvest credentials and deploy malware, revealing a persistent threat in Iran's cyber strategy.

Binarly has launched Transparency Platform 3.5, adding Java ecosystem support and enhanced cryptographic analysis to improve software supply chain security. The update includes enterprise YARA integration, new governance pipelines, and comprehensive risk reporting, enhancing visibility without needing source access.

State-sponsored actors are leveraging AI-powered malware, like PROMPTFLUX, to create dynamic and evasive cyber threats. North Korean and Iranian groups are particularly active in these sophisticated cyber operations, exploiting AI to bypass security measures and conduct data theft.

The SCIM framework is evolving to accommodate AI agents, enabling seamless identity management and security for non-human entities. SSOJet's unified console facilitates SCIM 2.0 readiness, allowing automated onboarding, auditing, and access control for AI agents, preparing enterprises for secure AI integrations.

Truffle Security has secured $25 million in Series B funding, led by Intel Capital and Andreessen Horowitz, to expand its enterprise-grade secret detection tools. The funding will enhance TruffleHog Enterprise and launch a new GCP Analyze add-on, aiming to bolster security teams' ability to manage and remediate leaked secrets efficiently.

Flare, a Montreal-based cybersecurity firm, has secured $30 million in funding to enhance its threat exposure management and identity exposure management services. This investment will bolster product development, enable strategic acquisitions, and support global expansion, affirming investor trust in Flare's innovative approach to cybersecurity challenges.

HMGUARD, a new framework, leverages adaptive prompting and reasoning in multimodal models to detect harmful memes, showcased at NDSS 2025. It surpasses existing tools in both public datasets and real-world scenarios, highlighting the need for improved detection methods amid diverse meme strategies.

Ransomware groups are leveraging automation and advanced tooling to enhance their ransomware-as-a-service models, increasing speed and efficiency in their operations. With 80% of groups using AI, breakout times are down to 18 minutes, posing a significant challenge for defenders who must adopt comprehensive security strategies.

Cloudflare has removed Aisuru botnet domains from its Top Domains list after they skewed rankings with excessive DNS queries to Cloudflare's 1.1.1.1 resolver. This move aims to balance security and accuracy, while experts urge vigilance with .su domains, commonly used by Aisuru, to spot potential threats.

Malanta, a cybersecurity startup from Tel Aviv, has secured $10 million in seed funding to advance its pre-attack prevention technology aimed at thwarting AI-driven threats. The company's innovative platform identifies and disrupts potential cyber threats before they become active, distinguishing itself from traditional post-breach detection methods.

Barracuda Networks has introduced the AI-driven Barracuda Assistant within its BarracudaONE platform to enhance security operations for organizations and service providers. The tool promises to streamline workflows and improve threat response efficiency, supported by real-time insights from Barracuda’s global threat intelligence network.

Taiwan is ramping up its domestic drone production to counter China's growing drone capabilities and potential aggression. By 2028, Taiwan aims to produce up to 180,000 drones annually, leveraging U.S. strategies and insights from the Russia-Ukraine conflict to enhance its defense strategy.

Russian hacker group Sandworm, or APT44, hit Ukraine's education, government, and grain sectors with data-wiping attacks in mid-2025. These attacks highlight a strategic shift towards disrupting Ukraine's critical economic sectors, with offensive cyber operations linked to broader geopolitical aims.

ClickFix malware campaigns now use video tutorials and OS detection to trick victims into self-infection, highlighting a shift in social engineering tactics. Security experts advise against executing unfamiliar terminal commands, as future campaigns may bypass traditional endpoint defenses.

Cavalry Werewolf executed a sophisticated cyberattack on a Russian state entity using custom backdoors and trojanized utilities like WinRAR and Visual Studio Code. Security experts advise vigilance against third-party downloads and recommend thorough scanning of installations to mitigate such threats.

Enterprise data growth coupled with increased cloud adoption is overwhelming security measures, leading to a rise in data loss incidents, with insider risk being a major concern. A recent report highlights that 85% of companies faced data breaches, driven largely by human errors and AI-induced risks, while security teams struggle with tool sprawl and coordination challenges.

IP cameras are highly vulnerable IoT devices, with 40% of them possessing at least one security flaw, a study of 25,000 cameras reveals. The expansion of unmanaged xIoT devices complicates asset tracking, increasing security risks, especially in sectors like healthcare and retail.