{"slug": "attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries", "title": "Attacking AI-based reverse engineering tools using weaponized binaries", "summary": "U.S. Navy security researchers Brian Crawford, Justin Phillips, and Patrick McClure published two papers on attacking AI-based reverse engineering agents via prompt injection, testing their approach against a setup combining Cline, GhidraMCP, Ghidra, and Qwen3-8B by embedding attack prompts in a small C program. The work also notes a practical constraint: Ghidra truncates string variables longer than 2048 characters, so attack strings must be short enough to survive decompilation and still reach the LLM.", "body_md": "Security researchers from the U.S. Navy - Brian Crawford, Justin Phillips, and Patrick McClure - published two papers last month on a groundbreaking topic: attacking reverse engineering AI agents using prompt injection techniques.\n\nThe authors propose various ways to attack the reverse engineering flow of AI-based tools and successfully test their approach on a setup with Cline, GhidraMCP, Ghidra, and Qwen3-8B - injecting prompts using a small C program. Quite impressive.\n\nImportant insight: Ghidra truncates string variables longer than 2048 characters. So keep in mind that the attack string has to be short enough to survive decompilation and attack an LLM.\n\nEnjoy the read, and remember that every new code not only makes something faster or better, but also increases the attack surface. Happy hunting!\n\nMore details:\n\nInvestigating Detection and Obfuscation of Prompt Injection Attacks Against Software Reverse Engineering AI Agents [PDF]: [https://arxiv.org/abs/2605.30677](https://arxiv.org/abs/2605.30677)\n\nAutomatically Attacking Software Reverse Engineering AI Agents [PDF]: [https://arxiv.org/abs/2605.30667](https://arxiv.org/abs/2605.30667)", "url": "https://wpnews.pro/news/attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries", "canonical_source": "https://it4sec.substack.com/p/attacking-ai-based-reverse-engineering", "published_at": "2026-10-12 19:01:31+00:00", "updated_at": "2026-10-12 19:20:05.824216+00:00", "lang": "en", "topics": ["ai-safety", "ai-agents", "ai-research", "large-language-models", "ai-tools"], "entities": ["U.S. Navy", "Brian Crawford", "Justin Phillips", "Patrick McClure", "Cline", "GhidraMCP", "Ghidra", "Qwen3-8B"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries", "markdown": "https://wpnews.pro/news/attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries.md", "text": "https://wpnews.pro/news/attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries.txt", "jsonld": "https://wpnews.pro/news/attacking-ai-based-reverse-engineering-tools-using-weaponized-binaries.jsonld"}}