EP24 The Glupteba Takedown: What Happens When Botnet Operators Show Up in Court with Pierre-Marc Bureau

In this episode of Behind the Binary, we sit down with Mac security researcher Christopher Lopez (L0psec). to discuss the evolving landscape of macOS malware. As Mac adoption continues to grow in enterprise environments and among developers, threat actors are increasingly targeting the platform. Chris breaks down current attack vectors, starting with the rise of social engineering tactics like ClickFix, fake software downloads, and malvertising campaigns. They explore how infostealer families like Atomic Stealer (AMOS) and persistent backdoor threats have matured, transitioning from simple cleartext scripts to complex, multi-layered obfuscation, C2 fail-safes, and anti-analysis checks. The conversation also covers DPRK APT campaigns targeting developers through fake job interviews and compromised VS Code environments. Finally, Josh and Chris discuss the impact of AI and LLM-driven workflows on the reverse engineering discipline. They dive into the balance between automated code deobfuscation, rapid detection engineering, and the loss of traditional manual reversing workflows as the industry shifts toward reversing at scale. Key Topics Covered:Enterprise macOS adoption and the expanding threat surface Initial access vectors: ClickFix, malvertising, and Terminal social engineering Evolution of Mac InfoStealers (AMOS/Atomic Stealer, MacSync) and persistence Advanced threats: DPRK developer targeting via VS Code tasks Obfuscation trends: Dynamic AES encryption, C2 key retrieval, and external fail-safe C2s AI's dual role in malware development and reverse engineering workflows System protections: TCC, SIP, Endpoint Security (ES) APIs, and macOS security prompts Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

"Throwing more LLM prompts at a packed binary won’t make your analysis faster. But combining code similarity with AI will."In Episode 29 of Behind the Binary, Jonas Wagner and Endre Bangerter of ThreatRay join the podcast to lay out the real-world state of binary similarity in the LLM era. They demystify the current AI hype cycle, explaining why large language models are structurally unsuited for raw, unguided binary analysis, and how traditional binary similarity techniques are finding a massive resurgence as the ultimate intelligence pre-filter.Learn how they automate the classification of complex compiled samples by separating known-good libraries from unique malware code, and how these combined workflows are shrinking threat attribution timelines down from days to minutes.THE SESSION:Beyond BinDiff: Tracing the transition from expert-crafted syntactic tools to deep semantic vector matching.The Token-Saver Pipeline: How to stop wasting LLM context windows on standard library runtimes (like Go and Rust) and target only custom, unique malware functions.Dynamic Unpacking: Leveraging sandbox execution and memory dumps to capture clean binary representations before static analysis begins.Anti-AI Obfuscation: The emerging landscape of adversarial techniques designed to throw off both similarity algorithms and LLMs.Scale and Attribution: Building indexed code repositories to search back in time and attribute code reuse across decades of malware families.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

In this episode, we're joined by security researcher Olivia Gallucci to explore macOS low-level security, kernel exploitation, and macOS threat trends. We break down how macOS internals differ from Linux—focusing on IOKit, C++ dynamic class resolution, and Vtable hijacking in kernel drivers. Olivia explains the mechanics behind use-after-free (UAF) vulnerabilities, heap shaping, and legacy kernel exploits, as well as modern Apple hardware and OS defenses like Pointer Authentication Codes (PAC) and kalloc type isolation (PAC IA/DA semantics). We also discuss the evolution of macOS threat research, the rise of macOS info stealers driven by cryptocurrency targeting, local detection telemetry using Endpoint Security (ES) and tools like Mac Monitor, and the current dynamics of vulnerability research and bug bounty programs. Key Discussion Points:Mac vs. Linux Internals: Why macOS isn't "just Linux," and how IOKit’s C++ driver framework creates unique attack surface. Vtable Hijacking & Kernel Exploits: How dynamic method resolution and C++ polymorphism are targeted using UAF, out-of-bounds writes, and heap shaping. Modern Defenses: How Apple leverages Pointer Authentication Codes (PAC IA/DA) and type-isolated heap allocators to disrupt traditional exploit primitives. The Rise of macOS Infostealers: What drove the surge in macOS stealer malware, their simple architecture, and social engineering delivery tactics. Detection & Telemetry: Monitoring local exploit failures, kernel panics, and process events via Endpoint Security (ES Logger) and open-source tools like Mac Monitor. Vulnerability Research Ecosystem: The role of AI in technical research, shifting bug bounty payouts, and the dynamics between vendor programs and third-party research. Resources mentioned:Olivia Gallucci's Blog: oliviagallucci.comMac Monitor by Brandon Dalton: https://github.com/Brandon7CC/mac-monitorJoin the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

When malware analysis first emerged, reverse engineers typically operated in a straightforward C landscape — standard Win32 APIs, no massive runtimes, and clear code paths. Today, the landscape has fundamentally shifted. The natural progression to C++ introduced object-oriented hurdles like virtual function tables and standard template library bloat, laying the groundwork for the complexities we see in today's modern ecosystem. In this episode of Behind the Binary, we sit down with Jae Young Kim from the FLARE team to discuss navigating this evolution and what it actually takes to reverse engineer modern compiled languages like Go and Rust.Jae walks us through his journey of diving head-first into the Go compiler architecture and source code to build a comprehensive internals reference for the FLARE Learning Hub. We discuss why understanding language quirks from the inside out is essential for professional analysts, look at the reality of recovering symbols from "stripped" Go binaries, and explore why traditional workflows like Time Travel Debugging (TTD) often get buried under runtime cruft. Finally, we look at the immediate future of the trade: how LLMs are already excelling at high-level decompilation and what languages he thinks are going to play a more significant role (hint - .NET AOT and Nim came up).THE SESSION:The Modern Language Shift: How the transition from C and C++ to sprawling, modern compiled runtimes has altered the day-to-day workflow of threat researchers and steepened the learning curve for beginners.Demystifying Go Internals: Jae’s perspective on why Go's compact design philosophy can actually make it a highly systematic, and occasionally pleasant, reversing experience once you master its unique quirks and calling conventions.The Myth of the "Stripped" Go Binary: A look at how tools like goReSym leverage intact compilation metadata to programmatically restore original function names and type definitions.AI and the Future Decompiler: An honest assessment of how current LLMs handle assembly-to-source reconstruction, where hallucination verification still requires a human in the loop, and how automated tooling will reshape the analysis pipeline.Emerging Blind Spots: A primer on the next operational hurdles for threat researchers, including the rising frequency of native ahead-of-time (.NET AOT) compilations and wildcard languages like Nim.References:FLARE Learning HubJoin the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

Mobile security boundaries rely on isolating remote, untrusted inputs from highly privileged system components. However, when new automated features are introduced, the available attack surface can shift—sometimes exposing unexpected code paths to remote attackers.In the latest episode of Behind the Binary, we sit down with Seth Jenkins from Google Project Zero to dissect a full two-bug, zero-click exploitation chain targeting the Pixel 9. By chaining a user-space decoder flaw with a kernel driver race condition and a kernel ASLR bypass, researchers achieved remote code execution and a device-wide SELinux sandbox escape.Key takeaways from our technical breakdown:The AI Attack Surface Shift: How implementing automatic voice message transcription inadvertently exposed the Dolby audio decoder (EAC3) to remote inputs.Defeating Userland ASLR: The mechanics of an integer overflow in user space (CVE-2025-54957) and how a partial pointer overwrite bypassed ASLR.Rigging the Race Condition: A look inside the "Big Wave" kernel driver use-after-free (CVE-2025-36934), a single bug that allows an attacker to jump from the mediacodec sandbox to the kernel. .The Predictable KASLR Reality: How standard ARM64 linear mapping combined with static bootloader placements negated kernel randomization—allowing two arbitrary writes to escalate privileges to root with SELinux disabled.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

"TTD is a paradigm shift in the way you interact with the target... Potentially, five years from now, when we talk about debugging, we will just by default go to TTD."In this episode, we are joined by Xusheng Li, a debugger architect and reverse engineering expert, to explore the evolution of Time Travel Debugging (TTD). While traditional debugging has remained largely stagnant for decades, TTD introduces a novel new way to debug by recording and replaying execution traces with total precision. Xusheng takes us behind the scenes of how this technology solves the "granularity problem" in malware analysis—moving from a high-level API overview down to instruction-level "ground truth" without ever needing to re-run the program.We break down the engineering required to record billions of instructions into a manageable trace, the power of querying execution data like a searchable database, and how a "sealed" execution history is changing the workflow for both software developers, malware analysts, and vulnerability researchers.THE SESSION:The Deterministic Leap: How TTD avoids the overhead of recording every single instruction by focusing only on non-deterministic events—like file reads, CPU ID calls, and system inputs—allowing billions of cycles to be reconstructed from a fraction of the data.The Death of "Step-Over": Why the future of debugging lies in querying an execution database rather than manually stepping through code, enabling researchers to instantly find every moment an API was called or a specific memory address was modified.Solving the Granularity Problem: How a single trace file provides a "safety net" for analysis, allowing researchers to start with a broad triage of behavior and then use a "microscope" to dig into specific crypto functions or obfuscated payloads later.Data Flow vs. Code Flow: A look at the shift toward "concrete data flow analysis," where researchers focus on the movement of sensitive buffers and keys rather than getting lost in the mental overhead of complex instruction sets and registers.The Mystery of the i9 Crash: A real-world troubleshooting case where TTD was used to identify a hardware-level microcode bug in a modern CPU that would have been nearly impossible to pinpoint with traditional tools.The AI Connection: Why the "fixed world" of a TTD trace is the ideal training ground for LLM-assisted analysis, providing a secure, deterministic environment for AI to solve intermediate-level reverse engineering challenges.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

"I thought that we would never hear about these people after they were named. But what was a surprise is that they actually hired a lawyer in New York... and they were like, 'Yeah, we're going to be taking part in this trial."In this episode, we are joined by Pierre-Marc Bureau from Google’s Threat Intelligence Group (GTIG) to unpack the unprecedented takedown of the Glupteba botnet. Active since 2011, Glupteba infected roughly 1 million Windows devices before Google launched a coordinated technical and legal strike. Pierre-Marc takes us behind the scenes of an investigation that evolved from reverse engineering binaries to a surreal showdown in a New York civil court.We break down how a single hardcoded string unraveled a massive criminal enterprise, the mechanics of using the Bitcoin blockchain for resilient command and control, and the bizarre moment when Russian cybercriminals actually hired a US lawyer to fight back.THE SESSION:The Blockchain Fallback: How Glupteba operators hid AES-encrypted blobs inside Bitcoin transactions, creating an un-takable backup C2 infrastructure if their primary domains went down.The Fatal OpSec Flaw: How one mistake—leaving the string get.voltronwork.com in a Go module—allowed Google to connect the botnet to Russian developer shops and Delaware shell companies.The Corporate Cyber-Cartel: Why the group operated like a legitimate tech startup, openly selling end-to-end "services" like proxy networks and compromised Google and Facebook accounts on the open web.The Extortion Twist: The surreal courtroom drama where the malware operators tried to extort Google for $1 million per defendant in exchange for private keys—a move that ended with the judge sanctioning their lawyer for $250,000.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

In this episode, we are joined by Robert Wallace, Joseph Dobson, and Blas Kajusner to dissect the new "Hybrid Heist." The panel argues that the era of isolated crypto-theft is over; sophisticated actors are now targeting the Web2 layer—the frontends, the developer workstations, and the cloud infrastructure—to bypass the immutability of the chain itself.We also break down "Ether Hiding," a technique where attackers store malware payloads directly on the blockchain to create an unstoppable Command & Control (C2) infrastructure that cannot be taken down by traditional authorities.THE SESSION:Immutable C2 (Ether Hiding): How threat actors are updating smart contract state variables to serve second-stage malware payloads, effectively turning the blockchain into a "dead drop resolver" that ignores domain blocks and takedown requests.The Hybrid Attack Surface: Why the massive Bybit heist wasn't a failure of cryptography, but a Web2 frontend attack on the "Safe Wallet" interface that tricked users into signing transactions they couldn't see.The "OpSec" Crisis: Why smart contract developers are the new "Domain Admins," and how simple phishing campaigns against personal devices are leading to nine-figure losses.The "Choke Point" Vulnerability: Why the decentralized ecosystem is still entirely dependent on centralized on-ramps and off-ramps, and how this dependency creates a "kill chain" that defenders can disrupt.Governance Attacks: The shift from exploiting code to exploiting consensus—how attackers are buying enough tokens to legally vote themselves the contents of a project's treasury.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

"Skilled adversaries have a 100% success rate against all of the defenses that we know about."In this episode, Kevin Harris defends that claim. We move past the standard "AI Safety" talking points to distinguish between the two attack vectors confusing the industry: Prompt Injection (an application-layer failure) vs. Jailbreaking ("gaslighting" the model via context shifting).Kevin argues that we haven't actually invented AI yet—we've just built a mirror that reflects our own intelligence (and psychosis) back at us. We also dissect the new model context protocol (MCP) and why giving "discretion" to agents that cannot think is potentially repeating the security mistakes of Web 2.0.THE SESSION:The "Pirate" Jailbreak: Why telling a model to be a pirate isn't just a party trick—it's a method of shifting the context window to bypass refusal patterns.The 100% Failure Rate: Why current defenses are only speed bumps for skilled adversaries, and why you are attacking the application, not the model."There Is No AI": Kevin’s theory on why LLMs are just "predictive text made 3 orders of magnitude better" and the danger of "AI-induced psychosis".The Agentic Threat (MCP): A deep dive into the model context protocol. Why client-side authorization is the new "Browser Security" battleground, and why we are handing "table saws" to users who don't know how to use them.The Fix: Why "Attention Functions" are the key to understanding (and securing) the future of these models.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube

In this episode, Dhillon Kannabhiran shares the gritty origin story of Hack in the Box (HITB), detailing how he dug a $20k financial hole to launch the first event in Malaysia before building it into a global brand.The conversation moves beyond conferences to explore the cutting edge of technology and creativity. Dhillon explains why "agentic" systems (like Xbow) signal the end of hand-built exploits and discusses the unique challenges of securing Web3 smart contracts. We also dive into the intersection of math and music, how AI tools like Suno are changing art, and why the "hacker ethos" applies to everything from bug bounties to content creation.Get the latest from FLARE's community efforts: Email [email protected] to join our mailing list for important announcements. Your information will not be shared and is used only for this purpose.Join the CommunityResearch Hub: Threat research, training events and news:https://cloud.google.com/security/flareThe FLARE Insider: Get community updates and announcements. To subscribe, email [email protected] THE SHOW:Subscribe: Apple Podcasts | Spotify | YouTube