Measuring Adversary Costs to Exploit Commercial Software...
summary
CITL measures how much work an attacker needs to newly exploit commercial binaries by scoring 100,000+ Windows, Linux, and OS X executables using static features and targeted fuzzing, and releases quantified “safety continuums.” Results use AFL fuzzing and Capstone disassembly, correlate exploit prices to hardening, and reveal surprises: Windows 10 images score consistently high while many macOS and Linux apps lack ASLR/stack guards, there are 7 million fortify opportunities and several distributions ship Anaconda compiled with 2005 GCC, making those bundles trivial targets.
AI-written summary · about
chapters (18)
- 0:00Start
- 2:35Certifications and Evaluations
- 6:43Os x Histogram
- 7:19Most Popular Browsers
- 8:42Office Suites
- 10:39Windows 10
- 14:38Complexity Measures
- 15:06Application Armoring
- 15:58Developer Hygiene
- 17:27The Dynamic Crash Testing
- 23:40File Code and Data Segment Sizes
- 24:41Main Report
- 25:33The Three Most Popular Browsers on Os X
- 28:15Source Code Fortification
- 28:50What Can We Learn about Linux and Source Code Fortification
- 34:42Development and Build Environments
- 41:03What Is Anaconda
- 46:51Timeline
related talks
23mSecurity Outgunned: Measuring Software Defined Attack Surface
38mCatch Me If You Can. LOLBins and the Adversaries Who Abuse Them
01h 26mEmbracing an Adversarial Mindset for C++ Security
20mSponsored Demo: The True Cost of Building at Machine Speed
52mBuilding Community-Driven Security Analysis for Your .NET Software Supply Chain
39mThe AI Arms Race in Vulnerability Management, Who’s Winning?
28mAdversarial Intelligence: Redefining Application Security Through the Eyes of an Attacker
20mEvolving from patch management to risk mitigation