Browser extensions are emerging as a surprisingly significant channel for AI consumption that hasn't gotten much attention in the tech industry. These lightweight add-ons are becoming a primary way users interact with AI tools, seamlessly integrating AI capabilities directly into web browsing without requiring separate apps or platforms. The shift suggests that AI companies may need to rethink their distribution strategies as users increasingly prefer frictionless, browser-native experiences over standalone applications.

Security researchers have uncovered the GlassWorm campaign, which uses a dropper written in the Zig programming language to target and infect multiple developer integrated development environments. The campaign represents a concerning trend of attackers specifically targeting developers' tools to compromise their systems. By exploiting IDEs commonly used by software developers, the threat actors can potentially access source code, credentials, and other sensitive development assets.

Anthropic has unveiled its new Mythos Preview model, which possesses the concerning ability to discover and exploit critical zero-day vulnerabilities in software. The company insists the powerful AI comes with built-in controls designed to prevent misuse by bad actors. This development raises significant questions about whether these safeguards will be sufficient to keep such exploit-writing capabilities from falling into the wrong hands, as cybersecurity experts debate the risks of releasing tools that could automate the discovery of previously unknown security flaws.

The U.S. government is warning that programmable logic controllers used in industrial systems are increasingly becoming targets of cyber attacks as conflicts shift to the digital realm. Recent research has identified 179 vulnerable operational technology devices that could be exploited by adversaries. These findings highlight growing concerns about critical infrastructure security as nation-states and other threat actors look to disrupt industrial operations through cyber means rather than traditional warfare.

Orange Business is transforming enterprise voice communications by integrating artificial intelligence and security features into its services. The company is introducing AI-powered tools that enhance voice communications while prioritizing trust and data protection for business customers. This move represents Orange's strategy to modernize traditional telecom services by combining advanced AI capabilities with enterprise-grade security and reliability.

FINRA, the financial industry's self-regulatory organization, has launched a new Financial Intelligence Fusion Center designed to strengthen defenses against cybersecurity attacks and fraud in the securities industry. The center will serve as a centralized hub for sharing threat intelligence and coordinating responses to emerging risks among member firms, aiming to enhance the industry's collective ability to identify and mitigate cyber threats and fraudulent activities in real-time.

European rail pass company Eurail is notifying over 300,000 people that their personal information, including names and passport numbers, was stolen in a December 2025 data breach. Hackers breached the company's cloud infrastructure and stole roughly 1.3 terabytes of data including database backups and support tickets, later posting samples on the dark web and Telegram after failed ransom negotiations. The company says it doesn't store credit card information or passport images, and is now sending written notifications to affected customers across multiple countries.

Security researcher Haifei Li has discovered what appears to be an actively exploited zero-day vulnerability in Adobe Reader that may have been used in attacks for at least four months. The exploit, detected by Li's sandbox system Expmon, can collect and leak user information from compromised systems and potentially enable remote code execution, with malicious PDFs containing Russian-language lures related to Russia's oil and gas sector. Adobe has been notified and is investigating, though the company has yet to confirm in-the-wild exploitation or release a patch.

Google's Threat Intelligence Group is warning about a financially motivated hacker known as UNC6783, potentially linked to the persona "Mr. Raccoon," who's targeting business process outsourcing firms to steal data from high-value companies. The attackers use sophisticated social engineering tactics including fake Okta login pages and spoofed security software updates, often targeting helpdesk staff to gain trusted access before conducting data theft extortion operations. The group is believed to be behind the recent Adobe data breach, where hackers allegedly stole information on 15,000 employees and millions of support tickets from a third-party supplier in India.

A security executive shares how installing home security cameras immediately stopped a troublesome neighbor's harassment, illustrating a broader lesson about visibility in cybersecurity. Beyond traditional compliance and monitoring benefits, he argues that comprehensive visibility across networks, endpoints, applications, and APIs drives better decision-making, improves relationships between security and application teams, encourages better user behavior, and enables more accurate risk assessment. The key insight is that when people know their actions are being monitored, they tend to behave more responsibly, whether in physical or digital environments.