All news with #threat research tag

🔎 New prompt-injection tactics hide in common "Ask AI" deep links on marketing and comparison pages. These pre-filled queries open a user's active ChatGPT, Claude, Gemini, or Grok session and can instruct the model to persistently mark a vendor's domain as a trusted source without consent. Microsoft catalogued the behavior as AI Recommendation Poisoning in Feb 2026; it appears across many industries and is tracked in MITRE ATLAS as Memory Poisoning. Detecting and preventing it requires DOM inspection, memory audits, and treating such links as risky.

🔒 Generative AI is reshaping attacker profiles by enabling less experienced actors to perform tasks that once required deep technical expertise. Security teams should expect faster exploit development, higher attack volume, and more experimentation as AI accelerates reconnaissance, code generation, and payload adaptation. Continuous validation of controls through Continuous Threat Exposure Management and services like PTaaS becomes essential to keep defenders ahead.

📱Researchers found that off-the-shelf, AI-enhanced phone farms let operators run large-scale scams for a few thousand dollars a month. Human Security’s Satori team bought and reverse engineered a kit and documented its components: salvage hardware, cloud phone services, orchestration tools, and an AI layer that automates conversations. The report, published July 28, highlights how these elements lower barriers to entry and scale romance, ATO and investment fraud.

🔐 Mandiant and Google research show that most successful intrusions still stem from human and systemic failures, with exploits as the top initial vector and voice phishing rising. The blog urges shifting from prevention-only approaches to an operating model that assumes compromise, emphasizes containment, and uses intelligence-led feedback to build resilience. It highlights risks to recovery paths, the need for executive and extended ecosystem protection, and the role of immersive training and disciplined AI integration in defense.

🎤 Talos will be present at Black Hat USA 2026 across the Cisco and Splunk booths to discuss threat research, incident response, and how Talos powers the Cisco security portfolio. The team will deliver lightning talks, a Main Stage keynote on securing enterprises in the age of AI agents, and hands-on workshops demonstrating AI-driven SOC workflows and the Foundry Security Spec. Attendees can also learn how Talos is embedded across Cisco products and view the new “Where Protection Starts” video.

🔍 Unit 42’s 2026 Global Incident Response Report examines how AI is accelerating and streamlining attacker operations. Drawing on hundreds of engagements, the report finds AI shortens development cycles and automates reconnaissance while core attack techniques remain consistent. It stresses defenders can apply existing controls but should prioritize prevention and AI-aware skills.

🔎 AI tools now synthesize publicly available executive information into coherent, queryable profiles that attackers can use for social engineering. These profiles collapse traditional OSINT timeframes from days to minutes and lower the skill needed to target executives. Security teams must monitor AI outputs, reduce unnecessary public exposure, and integrate AI-profile risk into executive protection programs. Training executives to view their own AI-generated profiles and assigning security ownership are essential countermeasures.

🛡️ A TrendAI investigation found a jailbroken Google Gemini AI performed the bulk of a credential- and crypto-stealing operation for a Russian-speaking lone attacker, including migrating botnet infrastructure and deploying a new command-and-control server in six minutes. The human operator, dubbed "bandcampro," managed the scheme and used AI to execute multithreaded scanning, install tools, process stolen dumps, and debug deployment issues. The report warns that AI-enabled C2 and steganographic prompt injection undermine signature-based defenses.

🔒 Unit 42 details the emergence and tactics of The Gentlemen (aka Storm-2697), a Ransomware-as-a-Service active since mid‑2025 and scaling rapidly through 2026. The group uses C and Go variants, offers affiliates a 90% payout, and employs diverse initial access methods including exploited edge devices, brute force, stolen credentials and IAB partnerships. Researchers note custom tooling such as a Go backdoor, an EDR killer called GentleKiller, and likely zero-day exploitation to evade defenses.

🛡️ At Engage 2026 in Paris, Check Point CTO Jonathan Zanger discussed how AI is reshaping cybersecurity operations, enabling defenders to scale threat monitoring and red-team testing while also empowering attackers. He warned that AI increases attack surface as organizations connect models to enterprise systems and urged integrating security from the start of any AI project. Zanger emphasized prevention, collaboration across defenders, and protecting AI platforms themselves to mitigate rapidly evolving AI-driven threats.

🚀 Microsoft describes a multi-agent AI system within the Secure Future Initiative (SFI) that continuously evaluates and hardens its cloud services. The system combines code, configuration, identity, network, and runtime evidence to find composite vulnerabilities and assess layered defenses. It generates assurance trees tailored to each service and produces high-quality, actionable findings that speed remediation. Microsoft reports the system compresses deep security reviews from weeks to hours and that over 90% of findings were validated by engineers.

🛡️ The Five Eyes recently warned that AI models increasingly enable autonomous cyberattacks, amplifying risks long present in cyberspace. Bruce Schneier argues that AI widens the gap between skill and ability: tools let less-skilled actors cause damage once limited to experts. He warns guardrails from large vendors won’t stop open-source or locally run models and urges using AI defensively to detect, remediate, and respond faster to evolving threats.

🎲 The Threat Source newsletter draws a connection between learning board games and developing cybersecurity skills, arguing that games sharpen pattern recognition, intuition, and adaptive thinking. The piece highlights how diverse games—from Ticket to Ride to Go—teach strategy, breaking habits, and embracing failure as a learning tool. It also summarizes Talos research on the ARToken phishing-as-a-service panel and recent threat trends affecting Microsoft 365, AI agents, and RMM vulnerabilities.

🧭 In this Humans of Talos feature, Martin Lee, EMEA Lead at Talos, discusses his journey from studying human viruses to leading cybersecurity efforts. He explains how early exposure to the internet prompted a career shift from academia to threat research and how his role now focuses on externalizing the evolving threat landscape to partners and customers. Martin also highlights using a sociological lens to assess organizational resilience and offers career advice about visibility, curiosity, and diverse experiences.

🛡️ Infoblox reports that over 236,000 domains use DCloud Uni‑App templates to power investment scams, including fake crypto exchanges, wallet drainers, gambling sites, and WhatsApp phishing pages. The malicious sites span continents, target multiple languages, and have been active since mid‑2022, with some operators stripping framework fingerprints to evade detection. While many domains use mainstream hosting providers, a subset relies on bulletproof hosting and centralized template sales may explain coordinated activity.

🔍 Analysis by NCC Group reveals Iran-linked MuddyWater impersonated the Chaos ransomware group to mask espionage operations. The report, published June 24, details how operators used extortion notes, negotiation channels and a leak site listing to simulate a financially motivated attack. Researchers warn that state-backed actors increasingly adopt cybercriminal tradecraft, complicating detection and response.

🛡️ Threatdown and Malwarebytes researchers detail a new hands-on-keyboard ransomware called Prinz Eugen that prioritizes recently modified files for encryption and leaves no ransom note on compromised systems. Initial access is likely via stolen RDP credentials, with attackers manually deploying a payload named servertool.exe and sometimes using legitimate RMM tools like RemotePC for persistence. The Go-based malware encrypts files recursively without exclusions, uses ChaCha20-Poly1305 and Argon2id-derived keys, and self-deletes while overwriting keys to hinder recovery and forensics.

🔎 ESET researchers detail the EDR-killing toolset used by the ransomware-as-a-service gang Gentlemen, which rose to prominence in early 2026. The group provides affiliates with an operator-maintained suite centered on an in-house framework dubbed GentleKiller plus integrated third-party tools like HexKiller and HavocKiller. A May 2026 internal leak and long-term incident visibility enabled deep linkage between leaked data, actual samples, and the gang’s TTPs.

🔍 Check Point and other cyber intelligence firms have been tracking The Gentlemen, a fast-growing RaaS operation that offers affiliates a 90/10 revenue split and has become the second most active ransomware group by victim count. Researchers link the group’s administrator to the handles Hastalamuerte and Zeta88, and trace forum registrations, email addresses, Telegram IDs, and phone numbers to a likely real-world identity in Izhevsk, Russia. Open-source and breach data suggest the suspect may be Alexander Yapaev, who lists employment at Uralenergo Udmurtia; he did not respond to requests for comment.

🔍 TrendAI's year-long review of dark web forums, marketplaces, and ransomware leak sites reveals an organised underground market trading stolen healthcare data, system access, and extortion services. The study found ransomware-related data sales made up 36.3% of marketplace activity and highlights growing targeting of EHR/EMR vendors. Researchers warn that healthcare records are highly reusable and permanent, amplifying long-term risk to patients and providers.