Hello, stay ahead with CISO Brief 🚀

🛡️ WatchGuard and ESET report two active campaigns spreading Windows and Android banking trojans across Latin America and Europe. The Grandoreiro campaign leverages DLL side-loading, WebRTC/STUN/ICE communications, and phishing to target Portuguese banks and international financial services. ESET details BTMOB, a rapidly evolving Android RAT sold as a service with an APK builder that enables mass phishing-based distribution and remote device control.

📧 Recent phishing campaigns exploit Google’s AppSheet platform to send convincing emails from a legitimate noreply{@}appsheet.com address, making them likely to bypass filters. Attackers craft personalized messages — urgent warnings or enticing job offers — to trick victims into submitting identity details on clone sites, then harvest credentials and data. The compromises can lead to account takeover, device control, and secondary targeted attacks using the stolen information.

🛡️ Researchers uncovered a malicious npm package named mouse5212-super-formatter that exfiltrates files from the /mnt/user-data directory used by Anthropic's Claude AI. OX Security describes the campaign, codenamed Malware-Slop, as a postinstall script that authenticates to GitHub using environment or hard-coded tokens, creates or targets a repository, and uploads local files to an attacker-controlled account. The package has been downloaded hundreds of times, and the linked GitHub account—created shortly before the package appeared—has since disappeared. Analysts noted the actor leaked a private token, suggesting poor OPSEC and possibly AI-assisted malware creation.

🔒 A single malformed character in a web request can allow unauthenticated attackers to bypass access controls in applications built on Starlette, the Python framework behind FastAPI. X41 D‑Sec disclosed the vulnerability (CVE‑2026‑48710) after finding it in a source‑code audit; Starlette’s maintainer released a patch via GitHub. The flaw stems from inconsistent parsing of the Host header when rebuilding request addresses, causing middleware to see a different path than the router. Researchers warn many model‑serving and AI infrastructure components are exposed unless a compliant reverse proxy rejects malformed Host headers.

📰 The FBI's 2025 Internet Crime Report has been published, offering a range of statistics and findings on cybercrime trends. The author notes they only recently became aware of the report and references associated press releases and news articles. The post, dated May 27, 2026, points readers to additional coverage and commentary on the report's contents.

🛡️ Cisco Talos disclosed four heap-based buffer overflow vulnerabilities in the MediaArea MediaInfoLib (v26.01) library, all of which can lead to arbitrary code execution when processing a malicious media file. The issues were found by Dimitrios Tatsis of Talos and have been patched by the vendor per Cisco’s third-party disclosure policy. Users can obtain Snort rules to detect exploitation and consult Talos for vulnerability advisories. Administrators should update MediaInfoLib to the vendor-released fixed versions promptly.

🛡️ Cybermindz warns that burnout among cyber professionals should be treated as a measurable operational risk rather than only a wellness concern. Their survey of 101 practitioners found frequent burnout and high emotional exhaustion, while their iRest® training study across 275 participants showed improved sleep, reduced exhaustion and lower attrition risk. Founder Peter Coroneos argues a risk-based framing can secure resources and support resilience.

🛡️ CrowdStrike, Google, and The Shadowserver Foundation coordinated to disrupt the Glassworm botnet by simultaneously takedown of four resilient C2 channels. The threat abused Solana blockchain memo fields, the BitTorrent DHT, Google Calendar events, and traditional VPS-hosted servers to persist and evade mitigation. Active campaigns targeted developers via malicious OpenVSX and VS Code extensions and later poisoned GitHub and npm artifacts. Infected hosts now beacon to a CrowdStrike-controlled IP and YARA rules have been published to detect compromise.

🔒 Cisco researchers warn that safety guardrails in several leading large language models (LLMs) can be bypassed through multi-turn conversations. They tested frontier models including ChatGPT, Claude, Gemini, Nova and Grok, finding many were susceptible to manipulation that yields disallowed outputs. Techniques such as roleplay, ambiguity, reframing, and persona adoption were effective, and model configuration affected resilience.

🔒 Google announces AI Threat Defense, an integrated, automated security system that uses Gemini, Mandiant, Wiz, and CodeMender to detect, prioritize, and remediate AI-powered threats. The platform combines multi-model scanning, live exposure mapping, and AI agents to validate exploitability, generate fixes, and accelerate remediation. It emphasizes machine-speed monitoring, autonomous response, and consolidated visibility across development and runtime environments to reduce attack surface and speed patching.

🛡️ CrowdStrike, together with Google and the Shadowserver Foundation, announced the simultaneous disruption of all command-and-control channels used by GlassWorm, a persistent campaign that has targeted software developers since early 2025. The operators trojanized VS Code extensions and poisoned npm and Python packages to deliver a data-theft framework capable of credential harvesting and system profiling. Multiple resilient C2 resolution layers were used — Solana memo fields, BitTorrent DHT, Google Calendar events, and commercial VPS hosts — all of which were neutralized in the coordinated action. CrowdStrike attributes the activity to likely Russia-based cybercriminals and warns about the severe risk posed by supply chain compromises to developer ecosystems.

🎯 Group-IB has identified over 4,300 fraudulent domains impersonating FIFA since last August, organized across six schemes and four threat actors targeting 2026 World Cup fans. The main operator, dubbed Ghost Stadium, uses a Chinese-speaking developer and a phishing kit that clones fifa.com, including its PingIdentity SSO flow, and leverages paid Facebook ads. Other actors include domain squatters, a PhaaS supplier and infostealer campaigns, which have already harvested around 2,500 FIFA credentials. Group-IB warns ticket fraud losses could reach into the hundreds of millions and advises fans to buy only from fifa.com, avoid crypto-based offers and enable MFA.

🔒 Researchers disclosed a vulnerability in Gitea that allowed unauthenticated remote attackers to pull private container images from affected deployments without credentials. Tracked as CVE-2026-27771, the issue affects all Gitea versions prior to 1.26.2, which contains the fix. Noscope estimates more than 30,000 deployments globally may be impacted, spanning healthcare, aerospace, retail, and ISPs. Users are advised to update to 1.26.2 or enable REQUIRE_SIGNIN_VIEW as a temporary mitigation.

🔒 The U.S. Cybersecurity and Infrastructure Security Agency (CISA) has ordered federal agencies to patch a critical, actively exploited privilege escalation flaw in the LiteSpeed cPanel user-end plugin, tracked as CVE-2026-48172. LiteSpeed released urgent updates to fix the issue in the lsws.redisAble function and advised administrators to check logs and block suspicious IPs. CISA added the flaw to its known exploited vulnerabilities catalog and required patches by May 29 under BOD 22-01.

🔍 EvidenceForge is an open-source project from Cisco Talos that generates correlated, multi-source synthetic security logs using a single canonical event model, causal ordering, and realistic background noise. It outputs synchronized telemetry across 20+ log formats (Windows, Linux, network, and EDR) from a version-controllable YAML scenario file and includes AI-assisted scenario authoring. The tool emphasizes deterministic generation, sensor-aware visibility, and built-in validation to produce datasets suitable for training, testing, and exercises.

🔒 More than two-thirds of UK businesses plan to increase cybersecurity spending over the next 12 months as AI adoption and geopolitical uncertainty reshape budgets. The Q1 2026 Barclays Business Prosperity Index found 68% of leaders expect higher cyber investment and 46% say new technologies raise their exposure. Large firms have led the increase, with average cyber spend hitting £505,000 so far in 2026, and cloud, cyber and AI account for 44% of planned tech budgets.

🔒 The Dutch National Police arrested a 35-year-old man from Buren suspected of multiple unlawful intrusions into AFC Ajax's computer systems earlier this year. The intrusions allowed access to data belonging to a few hundred individuals, modification of fewer than 20 stadium bans, and reassignment of purchased tickets. Ajax patched the exploited vulnerabilities, reported the breach to the Dutch Data Protection Authority and police, and the investigation remains ongoing.

🛡️ Recent developments show AI moving from automation to autonomous cyber operations, shifting how offense and defense interact. The Anthropic Mythos Preview and related incidents illustrate models discovering and chaining vulnerabilities with limited human direction, prompting coordinated defensive responses from major vendors. Policy and procurement are adapting, and security leaders must treat AI agents as principals, invest in adaptive defenses, and reframe risk models for continuous compromise.