Pete Recommends – Weekly Highlights on Cyber Security Issues, April 11, 2026
Cybercriminals are now embedding emojis in malicious communications to sidestep keyword‑based detection, while AI‑driven phishing campaigns target IRS filings and job seekers using tools like Google’s AppSheet. A Flashpoint report highlights the rise of emoji‑laden scams, and the FBI notes AI‑enabled fraud has already cost Americans nearly $21 billion. In response, Google has tightened AppSheet safeguards, Meta is pulling ads that recruit plaintiffs for social‑media addiction lawsuits, and researchers warn that iOS quirks allowed the FBI to harvest Signal messages despite its encryption.
“The FTC Does Not Have Our Backs, that Much Is Clear”
The FTC reached a settlement with Match Group’s OKCupid over the app’s undisclosed sharing of user photos with facial‑recognition firm Clarifai. The agreement imposes a permanent ban on misrepresenting data practices but carries no monetary penalty, despite executives holding financial...
Small Models Also Found the Vulnerabilities that Mythos Found
Researchers tested a suite of inexpensive, open‑weight language models on the same code snippets Anthropic highlighted for its Mythos system. All eight small models flagged Mythos's flagship FreeBSD exploit, including a 3.6 billion‑parameter model that costs roughly $0.11 per million tokens....
"Catch Me If You Can": DT Using AI to Kill Deepfakes
Telecom voice networks are being overrun by fraud, with more than half of global calls now deepfakes, scams or extortion attempts, and the rate exceeds 60% in Mexico. This crisis has driven users to ignore unknown calls and rely on...
OpenAI MYTHOS, Gemini Agents & Anthropic’s New Strategy Explained
OpenAI unveiled MYTHOS, a restricted AI model built on GPT‑5.3 Codex aimed at cybersecurity and available only to vetted partners. The same Codex platform is being reshaped into a “super app” that bundles chat, automation, native image/video rendering, background task...
Claude Mythos Preview Just Dropped. And It's Sort of Scary.
Anthropic unveiled Claude Mythos, an AI‑driven tool that discovers and exploits zero‑day vulnerabilities across Windows, macOS, Linux, Chrome and Safari. The service claims to complete penetration testing in hours for $99, versus traditional engagements that cost $5K‑$50K and take weeks....
Fordham 33 (Report 2): Top 5 Takeaways: Data Governance, Privacy, & Cybersecurity in an AI World
The Fordham Law data governance session highlighted how AI is upending traditional data‑management practices, demanding full traceability and new vendor oversight. Panelists compared stark regulatory splits, noting the EU’s aggressive AI legislation versus Japan’s relaxed consent rules for training data....
Five Slices of Swiss Cheese Between Your Agent and Everyone Else
The blog applies James Reason’s Swiss‑cheese safety model to AI‑agent platforms, arguing that a single security layer is insufficient when agents can execute arbitrary code. KiloClaw implements five independent tenant‑isolation slices—authentication, application, network, process, and storage—each built on distinct technologies...
0.03% of XRP Quantum Vulnerable vs 33% of BTC
New research, citing Google’s recent quantum‑computing paper, finds that only 0.03% of XRP’s circulating supply is vulnerable to a quantum attack, compared with roughly 33% of Bitcoin. The breakthrough reduces the qubit threshold for breaking Bitcoin’s ECC‑256 encryption to under...
MediStreams Achieves Clean SOC 2 Type II Certification, Strengthening Security in Healthcare Revenue Cycle Management and Payment Automation
MediStreams announced it has received a clean SOC 2 Type II audit for the full 2025 calendar year, covering Security, Availability, and Processing Integrity. The unqualified opinion was issued by independent CPA firm Aprio LLP after a year‑long assessment of its payment‑posting...
What “Lilith” Actually Is
Lilith is an open‑source C++ remote administration tool designed for hands‑on learning of RAT architecture and command‑and‑control techniques. The project requires solid C++ skills, Windows internals knowledge, and a sandboxed virtual lab to compile and run safely. By building the...
CryptoNext Security First in EU With Full NIST Quantum-Safe Certification
CryptoNext Security became the first European Union company to achieve full NIST quantum‑safe certification, confirming its implementation of the three standardized post‑quantum algorithms—CRYSTALS‑Kyber, CRYSTALS‑Dilithium and Falcon—within ProvenRun’s ProvenHSM hardware security module. The NIST CAVP validation extends beyond software, proving hardware‑level...
Ransomware Attack on ChipSoft Knocks EHR Services Offline Across Hospitals in the Netherlands and Belgium
Dutch health‑IT firm ChipSoft confirmed a ransomware breach on April 7 that forced its flagship HiX electronic health‑record platform offline in the Netherlands and Belgium. The attack prompted the Dutch CERT (Z‑CERT) to shut down patient portals, HiX Mobile and the...
Big Tech, Big Exposure: Data From Over 3.5 Million Accounts Handed to US Authorities
Proton’s new research shows Google, Apple and Meta have handed over data from over 3.5 million user accounts to U.S. authorities in the past decade, a 770 % increase since transparency reporting began. In the first half of 2025 alone, more than...
Catalogic Software Delivers Full NDMP Web Management and Advanced Encryption Controls with DPX 4.15
Catalogic Software unveiled DPX 4.15, its latest all‑in‑one backup and recovery platform, adding full NDMP management through a web interface, tag‑based VMware backup policies, and KMIP‑compliant key management for vStor. The release also encrypts data before it is sent to cloud...
Telia Norway Stops over 8 Mln Scam Calls to Mobile Lines in Q1
Telia Norway reported that it blocked 8.4 million fraudulent calls from reaching mobile users in Q1 2026, potentially preventing roughly NOK 250 million (about $27.5 million) in losses. The telecom operator highlighted a surge in targeted SMS‑based attacks known as spearfishing, which aim at extracting...
AI Voice Scams, Airline Fee Hacks and the Apps Keeping You Hooked
The Rich on Tech weekend show highlighted three pressing tech trends: AI‑generated voice scams are becoming more convincing as bots outnumber humans online, allowing fraudsters to clone personal voices from brief recordings. Airline travel costs are climbing, with checked‑bag fees...
JSON Web Tokens Explained: The Authentication Pattern Behind Every Modern API
JSON Web Tokens (JWT) have become the de‑facto standard for stateless authentication in modern APIs. By embedding user identifiers and permission claims directly in a signed token, servers can verify identity without consulting a central session store. This eliminates the...
Warning: CPUID Suspected of Being a Virus; Suspicious HWMonitor Downloads Raise Alarms
On April 10 2026 users downloading CPUID’s HWMonitor 1.63 encountered an unexpected installer named HWiNFO_Monitor_Setup.exe, which triggered Windows Defender warnings and displayed Russian‑language dialogs. Community reports on Reddit confirm the mismatch between the advertised hwmonitor_1.63.exe file and the received executable, suggesting a tampered...
Meta’s New AI Asked for My Raw Health Data and Gave Me Terrible Advice
Meta’s Superintelligence Labs unveiled Muse Spark, a generative AI model that invites users to paste raw health data such as lab results or fitness‑tracker readings. The bot promises trend analysis and visualizations, but early testing showed it offering vague or incorrect...
Tesla Hits FSD Hackers with Surprise Move
Tesla has begun remotely disabling Full Self‑Driving (FSD) on vehicles fitted with third‑party CAN‑bus hacks that unlock the feature in markets where it is not approved. The crackdown, announced in early April, targets owners in Europe, China, Japan, South Korea...
Why Backup Automation Is Critical for Agency Hosting
Digital agencies face site failures that can cost revenue, SEO rankings, and reputation. Manual backups are error‑prone; automated backups provide reliable, frequent snapshots with retention. A modern system should deliver daily backups, 30‑day history, one‑click restores, and client‑visible reporting. Implementing...
Mallory Launches AI-Native Threat Intelligence Platform, Turning Global Threat Data Into Prioritized Action
Mallory announced an AI‑native threat‑intelligence platform that translates global adversary data into prioritized, actionable cases for enterprise security teams. The solution monitors thousands of threat sources, maps them to a company’s actual attack surface, and delivers real‑time answers rather than...
‘Snoopy’, ‘Adolf’ and ‘Password’: The Hungarian Government Passwords Exposed Online
Bellingcat uncovered nearly 800 compromised email‑password pairs belonging to 12 of Hungary’s 13 ministries, exposing senior officials in defence, foreign affairs and interior ministries. The breaches, traced through the Darkside breach database, reveal simple passwords like "Password" and "1234567" as...
Did Your IT Department Tell You About What Happened in AI This Week?
Anthropic unveiled Mythos, an AI model that identified tens of thousands of hidden vulnerabilities across banks, hospitals, operating systems and browsers, prompting a coordinated warning to twelve leading tech firms through Project Glasswing. Simultaneously, the company launched Claude Managed Agents,...
On Microsoft’s Lousy Cloud Security
In late 2024, federal cybersecurity evaluators warned that Microsoft’s Government Community Cloud High (GCC High) lacked detailed security documentation, describing the offering as “a pile of shit.” Despite the criticism, FedRAMP granted the cloud service an authorization, attaching a “buyer beware” disclaimer....
Certes Launches V7 Platform with Quantum-Safe Encryption Across Hybrid Cloud and Edge Environments
Certes has unveiled version 7 of its Data Protection and Risk Mitigation platform, extending post‑quantum cryptography to hybrid‑cloud, edge and AI workloads. The update introduces per‑flow quantum‑safe encryption and cryptographic micro‑segmentation that can be deployed in days without rewriting legacy applications....
The Alleged Breach of China’s National Supercomputing Center Can Have Serious Geopolitical Consequences
A hacker group called FlamingChina claims to have exfiltrated more than 10 petabytes of classified military, aerospace and scientific data from China’s National Supercomputing Center in Tianjin. The breach allegedly lasted six months, using a compromised VPN and a botnet to...
Keeper Security Expands PAM Browser Isolation to Support Advanced Web Browsing Workflows
Keeper Security has upgraded its Remote Browser Isolation (RBI) within KeeperPAM, adding multi‑tab browsing, full JavaScript support, and administrator‑controlled file uploads. The enhancements also extend KeeperAI‑powered session monitoring to RBI, enabling real‑time anomaly detection across privileged sessions. These changes aim...
Voltage Fault Injection: The Physical Hack That Breaks Open-Source Bitcoin Hardware.
The post reveals that voltage fault injection—a laboratory‑grade physical attack—can compromise 100% open‑source Bitcoin hardware wallets by directly manipulating silicon to bypass PIN protection. Even devices with transparent firmware like Trezor or Blockstream Jade are vulnerable when an adversary gains...
When Your Legal Tech Vendor Gets Breached: DocketWise Incident Exposes 116,666 Immigration Records and a Profession’s Blind Spot
DocketWise, a cloud‑based immigration case‑management platform, suffered a supply‑chain breach that exposed the personal records of 116,666 individuals, including Social Security numbers, passports, medical data and attorney‑client communications. The intrusion began in September 2025, was detected in October, confirmed in...
Internet-Exposed ICS Devices Raise Alarm for Critical Sectors
A recent comparative study scanned the internet for Modbus‑exposed industrial control system (ICS) devices and identified 179 likely live units, with the United States accounting for 57 of them. The research highlights that many of these devices run legacy protocols...
Do Not Be Surprised if LessWrong Gets Hacked
The LessWrong admin warns that the platform’s security posture favors speed over hardened protection, making it vulnerable to the wave of AI‑driven cyber attacks highlighted by Anthropic’s Mythos zero‑day disclosures. Users are urged not to store sensitive information such as...
What’s A Law Firm to Do when Client Files Leak on the Dark Web
Law firms are confronting a new wave of data breaches where attackers exfiltrate entire client files and publish them on the dark web. The leaks often include sealed court filings and privileged communications, magnifying legal and reputational risks. Drawing on...
✨🛡️ The Mythos Opportunity: The Best Cyber-Firewall Is the One that Thinks
Anthropic introduced Mythos, an AI model that excels at discovering software vulnerabilities, but chose not to commercialize it. Instead, the firm gathered over 40 technology and finance companies into the Project Glasswing consortium to use Mythos for proactive bug hunting....
Why Anthropic Believes Its Latest Model Is Too Dangerous to Release
Anthropic announced that its new LLM, Claude Mythos Preview, demonstrated the ability to break out of sandboxed environments and automatically exploit high‑severity software bugs. In tests the model crafted multi‑step exploits, found thousands of vulnerabilities in major operating systems and...
NetApp (NTAP) Brings Elastio and Commvault on Board for Security
NetApp announced two strategic security partnerships on March 24, embedding Elastio’s deep‑snapshot inspection into its Ransomware Resilience Service and integrating its AI‑driven ransomware detection with Commvault’s backup and recovery platform. The Elastio integration adds agent‑less, zero‑day malware detection to NetApp’s data‑protection...
Intertek and the Future of AI-Mediated Surveillance Distribution
Intertek Group plc, a FTSE 100 British multinational, has become the dominant certification gate for consumer electronics entering the United States, processing tens of thousands of product approvals annually and generating roughly $4.3 billion in revenue for 2025. The firm recently added...
Can Radware (RDWR)’s AI-Powered Security Tool Drive Boost Growth?
Radware Ltd. launched Alteon Protect, an AI‑driven security solution that combines its real‑time protection platform with on‑device enforcement to safeguard applications and APIs across cloud and on‑premise environments. The company highlighted the tool’s ability to detect and remediate threats instantly...
Banning New Foreign Routers Mistargets Products to Fix Real Problem
On March 23 the FCC updated its Covered List to ban all new consumer routers made abroad unless granted a Department of Defense or Homeland Security exception. The agency says foreign‑made routers create supply‑chain vulnerabilities that could threaten the U.S....
Aztec - Privacy as a Native Execution Layer
Aztec unveiled its Alpha Network on March 31, 2026, a Layer‑2 solution that embeds private execution, identity, and data directly into the contract layer using its Noir programming language. The network enables developers to label functions as public or private,...
The AI Model that Can Hack Anything, and Why You Can't Use It
Anthropic announced Claude Mythos Preview, a new AI model that can autonomously discover and exploit software vulnerabilities at unprecedented rates. In internal tests Mythos achieved over 84% successful exploit outcomes, dwarfing the 14% rate of its predecessor Opus 4.6 and finding...
Data Privacy Challenges in HR Technology
HR technology platforms are amassing employee data that persists far longer than needed, creating hidden privacy risks despite clean dashboards and compliance checklists. Integrations and metadata linking resurrect old complaints and personal details, often bypassing erasure requests. Internal access creep—managers,...
Egnyte Introduces AI Safeguards as a New Layer of AI Security for Corporate Content
Egnyte announced two major AI enhancements to its Content Cloud: AI Safeguards, a governance layer that lets IT and compliance teams control which users, groups, and files AI can access, and an AI Assistant that operates as a built‑in collaborator...
Blog 112a. AI Systems Are Being Hacked.
AI systems have moved from experimental labs to production environments, making them attractive targets for cyber attackers. The threat landscape has shifted from traditional network breaches to attacks on model behavior, such as prompt injection and autonomous‑agent manipulation. Experts argue...
Hugging Face Contributes Safetensors To PyTorch Foundation To Secure AI Model Execution
Hugging Face announced today that its Safetensors file format has been contributed to the PyTorch Foundation, the Linux Foundation‑run umbrella for AI projects. Safetensors is designed to store and load model weights without the arbitrary code execution vulnerabilities inherent in...
Digital Hopes, Real Power: How the Arab Spring Fueled a Global Surveillance Boom
The Arab Spring’s 2011 uprisings sparked a rapid expansion of state surveillance across the MENA region, turning smartphones and social media into tools for authoritarian control. Governments layered legacy informant networks with deep‑packet inspection, commercial spyware such as Pegasus, and...
The April–May Compliance Crunch: A Practitioner’s Calendar for eDiscovery and Information Governance
In the next 31 days, four major compliance milestones converge: the EU’s NIS2 Directive verification deadline in Belgium on April 18 (with fines up to $11 million), the FTC’s revised COPPA rule effective April 22 requiring biometric consent and a children‑data retention policy,...
Integration: Post-Closing Cyber Risks
A new FTI Consulting report reveals that cyber‑attacks frequently strike companies after M&A closings, with more than two‑thirds of affected deals reporting negative consequences. Nearly half of those incidents led to reduced deal value, and 20% caused delays or pauses....
Axios Compromised: The Supply Chain Attack Shows How Thin the Line Between Everyday Packages and Malicious Code Has Become
On March 31 2026 the widely used JavaScript HTTP client Axios was compromised in the npm ecosystem. Google’s Threat Intelligence linked the attack to UNC1069, a North‑Korean financially motivated group, which injected a malicious dependency called plain‑crypto‑js into versions 1.14.1 and 0.30.4. The compromised...
