Good Morning Security Gang,
Today’s show is one of those rare moments where multiple fault lines in cybersecurity crack at the same time. We’re not looking at isolated incidents we’re seeing systemic stress across endpoint security, AI tooling, patch management, and financial ecosystems. The common thread is clear: the controls we trust most are now being actively exploited.
Today’s episode highlights a convergence of risks across enterprise environments. Attackers are no longer focused on bypassing defenses—they are repurposing trusted systems as attack vectors. From Defender zero-days disabling detection capabilities, to AI frameworks introducing remote code execution risks by design, to crypto theft reaching industrial scale, the modern threat landscape is defined by speed, scale, and trust exploitation.
Organizations must shift from a prevention-first mindset to one centered on resilience, validation, and layered defense, particularly across identity, AI, and developer ecosystems.
Microsoft Defender Zero-Days – Security Tools as Attack Vectors
Microsoft issued emergency patches for three actively exploited zero-days in Defender for Endpoint. These vulnerabilities allowed attackers to tamper with detection mechanisms, effectively disabling security visibility.
This marks a critical shift. Endpoint Detection and Response (EDR) is no longer just a defensive layer, it has become part of the attack surface. If attackers can neutralize your detection stack, they gain time, persistence, and operational freedom.
The implication is significant: organizations relying on a single EDR solution are operating with a single point of failure.
“Attackers aren’t bypassing your defenses—they’re turning them into theirs.” James Azar and continues "That's why you need two endpoint products, not just one. It's a sad truth, but you've got to layer them. They might be able to silence Defender, but they can't silence SentinelOne or CrowdStrike. Attackers are no longer just evading endpoint security, they're actively weaponizing it."
Domain Controller Patch Failure – Operational Risk from Remediation
Simultaneously, Microsoft’s April patch cycle introduced instability in domain controllers, triggering reboot loops and widespread authentication failures.
This resulted in cascading operational disruptions, including login failures, VDI outages, and identity system degradation.
This is the paradox of modern security operations: patching is essential, but poorly validated patches can introduce systemic risk. Identity infrastructure must be treated as mission-critical systems requiring controlled deployment pipelines.
AI Supply Chain Exploitation – Marimo and Hugging Face
The Marimo RCE vulnerability, exploited within hours of disclosure, demonstrates how quickly attackers are operationalizing weaknesses in AI tooling.
Attack chains now include trusted platforms like Hugging Face, combined with decentralized command-and-control infrastructure, making mitigation more complex.
This signals a new reality: AI development environments are now enterprise attack surfaces, often deployed without the rigor applied to traditional infrastructure.
Anthropic MCP Vulnerability – Systemic AI Risk
A critical design flaw in Anthropic’s Model Context Protocol (MCP) introduces remote code execution risk across a wide ecosystem of AI development tools.
The scale of exposure millions of downloads and hundreds of thousands of instances—combined with the vendor’s stance that the behavior is “by design,” raises serious concerns about security maturity in AI frameworks.
This is not a bug. It is an architectural risk.
Lazarus Group Crypto Heist – Industrialized Financial Attacks
North Korea’s Lazarus Group executed a $290 million exploit against KelpDAO, leveraging weaknesses in cross-chain bridge architecture.
This follows a repeatable playbook: compromise infrastructure, manipulate trust assumptions, and extract value at scale.
Layer Zero is attributing the attack to North Korea’s Lazarus Group, specifically the TraderTraitor cluster.
The attack chain:
Compromised the downstream RPC endpoint that DVN relied on
Used DDoS to force failover to the poisoned endpoint
Spoofed cross-chain messages through KelpDAO’s single verifier setup
KelpDAO is disputing the narrative, arguing the single verifier configuration was Layer Zero’s own default rather than an unusual choice.
This correlates with Bybit’s $1.5 billion heist playbook and the broader DPRK revenue engine we’ve been tracking. April 2026 is now the worst month for crypto hacks since February 2025, with over $606 million lost in 18 days.
The volume of crypto theft in April alone underscores a critical trend financial cybercrime is now operating at nation-state scale and efficiency.
Salesforce Campaign – ShinyHunters Expands Targeting
ShinyHunters continues its campaign targeting Salesforce environments, now focusing on high-value organizations like Aman Resorts.
The Aman Resorts story is short but sharp, and the clock is ticking. Ransomware.live and Hookfish report Shiny Hunters added the ultra-luxury hospitality brand Amman Resorts to its leak portal on April 19th, claiming 500,000 Salesforce records of PII with an April 21st deadline to pay or see the data posted publicly.
This listing is the latest in Shiny Hunters’ active Salesforce-centric campaign that has also touched 7-Eleven, Pitney Bowes, Canada Life Assurance Company, and Marcus & Millichap in the last two weeks.
Rather than exploiting platform vulnerabilities, attackers are abusing integrations, OAuth scopes, and API access.
This reflects a broader shift toward identity and integration layer exploitation, where traditional perimeter defenses offer little protection.
BlueSky DDoS – Availability as a Primary Target
A multi-day DDoS attack against BlueSky disrupted core platform functionality, reinforcing that availability remains a critical component of security.
As organizations adopt decentralized architectures, mitigating volumetric and application-layer attacks becomes increasingly complex.
Internet Exposure – FTP as a Persistent Weakness
Despite years of awareness, over 2.4 million internet-facing FTP servers still operate without encryption. This is not a sophisticated threat, it is a failure of basic security hygiene at global scale.
Cleartext protocols continue to provide attackers with effortless access to credentials and sensitive data.
Key Action Items
Deploy layered endpoint security controls to mitigate EDR tampering risks
Implement staged patching and validation for identity infrastructure
Secure AI and data science environments with enterprise-grade controls
Audit and restrict AI agent frameworks and MCP integrations
Enforce multi-verifier models in blockchain and DeFi architectures
Review SaaS integrations, OAuth scopes, and API access patterns
Strengthen DDoS response strategies with pre-defined mitigation plans
Eliminate cleartext protocols and enforce encryption across all services
Monitor for credential theft and privilege escalation indicators
Continuously validate trust assumptions across all systems
"The pattern on today's show is brutal and consistent: your endpoint product is being weaponized against you, your patch pipeline is breaking the infrastructure it's supposed to protect, your AI tool chain from notebook servers to the MCP standard itself is a fresh RCE buffet, nation-state crews are draining DeFi bridges for a quarter billion at a time, Shiny Hunters is turning Salesforce connected apps into a breach factory, volumetric DDoS can still take major social platforms offline for days, and millions of unencrypted FTP servers still leak credentials across the public internet." James Azar
James Azar’s CISOs Take
What we’re seeing today is not a series of independent failures, it’s a systemic shift in how cyber risk manifests. Endpoint protection, AI tooling, patching processes, and financial systems are all being tested simultaneously. This isn’t coincidence. It’s the result of attackers identifying where trust has been overextended and exploiting it at scale.
The path forward requires a fundamental change in approach. Security can no longer rely on static controls or assumptions of safety. Every system must be treated as potentially compromised, every integration as a risk vector, and every layer as something that must be continuously validated. Organizations that embrace this mindset will adapt. Those that don’t will continue reacting to incidents rather than preventing impact.
Stay Cyber Safe
