Category: Black Hat

Black Hat USA 2025 | Reinventing Agentic AI Security With Architectural Controls

  • Haco
  • March 13, 2026
  • 4 min read

"When Guardrails Aren't Enough: Reinventing Agentic AI Security With Architectural Controls," delivered by David Brauchler III from NCC Group. The Core Thesis The central argument of the presentation is that guardrails are not security boundaries. Much like Web Application Firewalls (WAFs) in the early days of the internet, AI guardrails are merely statistical heuristics. They reduce risk but do not provide "hard" security guarantees and can always be bypassed by a determined attacker. As AI evolves into "agentic" systems—where models can execute tool calls, read databases, and take actions—relying solely…

Black Hat USA 2025 | Invoking Gemini for Workspace Agents with a Simple Google Calendar Invite

  • Haco
  • March 10, 2026
  • 3 min read

"Invitation is All You Need! TARA for Targeted Promptware Attack against Gemini-Powered Assistants," presented by Ben Nassi, Or Yair, and Stav Cohen. Core Premise The presentation highlights a new, highly practical class of cyberattack called "Promptware," specifically targeting Large Language Model (LLM) powered personal assistants like Google's Gemini for Workspace and Android. The researchers demonstrate how an attacker can completely compromise a user's AI assistant simply by sending them a Google Calendar invitation containing hidden, malicious instructions. The Attack Mechanism: Indirect Prompt Injection Unlike traditional hacking that targets memory corruption or…

Black Hat USA 2025 | Training Specialist Models: Automating Malware Development

  • Haco
  • March 4, 2026
  • 3 min read

"Training Specialist Models: Automating Malware Development" explores how small, specialized Large Language Models (LLMs) can be trained to outperform massive generalist models in specific, highly technical tasks—specifically, the creation of evasive malware. Here is a summary of the key points: The Problem with Current ModelsAvery identifies a gap in the current AI landscape for offensive security professionals: Large Generalists (OpenAI, Anthropic): These models are highly capable but come with privacy concerns, high costs, and strict safety filters (refusals) that make them difficult to automate for red teaming. Small Local Models…

Black Hat USA 2025 | Watching the Watchers: Exploring and Testing Defenses of Anti-Cheat Systems

  • Haco
  • March 3, 2026
  • 3 min read

Introduction to the Anti-Cheat Ecosystem The World of Game Cheats: The speakers explore the fast-paced, high-stakes battleground between cheat developers (attackers) and anti-cheat systems (defenders) in modern competitive shooter games []. The Cheat Economy: Cheating is a massive industry. Cheats are often sold via subscription models by well-run, sometimes legally registered companies, with some cheats costing upwards of $200 a month []. Because it is so lucrative, the attack-defense cycle is incredibly rapid. The Shift to the Kernel: Historically, cheats operated in user mode. As anti-cheats adapted, the…

Keynote: Perspectives on Trust in Hardware Supply Chains

  • Haco
  • September 30, 2025
  • 3 min read

This talk, titled "Keynote: Perspectives on Trust in Hardware Supply Chains" [] by Bunny Huang, discusses the complexities and vulnerabilities within hardware supply chains. Key points from the talk include: Diversification and Simplification: In chaotic times, it's beneficial to diversify by having multiple, hyper-efficient locations rather than centralized single points of failure. Simplifying business processes and legal contracts can reduce complexity and improve understanding for everyone involved []. Trust in Hardware: The speaker suggests moving the "root of trust" or "source of truth" into hardware, such as using…

Tinker Tailor LLM Spy: Investigate & Respond to Attacks on GenAI Chatbots

  • Haco
  • September 25, 2025
  • 3 min read

In the "Tinker Tailor LLM Spy: Investigate & Respond to Attacks on GenAI Chatbots" talk by Black Hat, Ellen Scott discusses the increasing ubiquity of Generative AI chatbots and the security incidents that can arise from their misuse. The talk outlines three main incident scenarios and provides a playbook for investigation and response []. Here's a summary of the key takeaways: Chatbot Risk Classification []: Low Risk: Chatbots providing general information (e.g., a weather chatbot). Incidents primarily involve brand damage, like a chatbot giving Taylor Swift-themed weather reports…

Microarchitecture Vulnerabilities: Past, Present, and Future

  • Haco
  • September 4, 2025
  • 1 min read

Past Present Logic Issues Exploitation Techniques Physical Domain in Software Mitigation Efforts Physical hardware cannot be changed in the field Vendors build in "Survivability features" Microcode is the most common used tool for mitigations Other firmware is also used "Chicken bits" to disable / change behavior Some issues are best mitigated in software Mitigations are not always possible/reasonable and almost difficult and time-consuming to engineer Prevention Pre-silicon Post-silicon Future Take Aways

SysBumps: Exploiting Speculative Execution in System Calls

  • Haco
  • September 2, 2025
  • 3 min read

The video presents an attack technique called SysBumps, demonstrated by researchers Hyerean Jang, Taehun Kim, and Youngjoo Shin at Black Hat Europe 2024. Here’s what it’s about: What SysBumps DoesSysBumps breaks Kernel Address Space Layout Randomization (KASLR) on macOS systems running on Apple Silicon, including M-series chips. It uses speculative execution within system calls, triggering side-channel behaviors that allow an unprivileged attacker to detect kernel memory layout. How the Attack WorksBy system calls that involve speculative execution, attackers can influence the translation lookaside buffer (TLB). This manipulation allows…

Bypassing ARM’s Memory Tagging Extension with a Side-Channel Attack

  • Haco
  • September 2, 2025
  • 2 min read

This explains a research project on how to bypass ARM's Memory Tagging Extension (MTE), a hardware feature designed to prevent memory corruption vulnerabilities. Here are the key takeaways: ARM MTE: MTE works like a "lock and key" system. Pointers have a "key" (a 4-bit tag) and memory objects have a "lock" (also a 4-bit tag). If the key and lock don't match when a pointer tries to access memory, the program will crash, preventing an attack. The Challenge: The tags are randomly generated, making it difficult for an…

Uncovering Supply Chain Attack with Code Genome Framework

  • Haco
  • September 2, 2025
  • 2 min read

This talk from IBM Research focuses on using AI and machine learning to combat supply chain attacks. The presenters highlight the increasing lack of trust in software due to major security breaches like the XZ backdoor. Here are the key takeaways: The Problem: There's a "semantic gap" between what code is expected to do and what it actually does. This gap is exploited in supply chain attacks where malicious code is hidden in software updates or open-source projects. The Solution: The researchers introduce the "Code Genome Framework," an…