Bugcrowd brings Mayhem AI to bear on ethical hacking community | Computer Weekly

"Mayhem - which won the Darpa Cyber Grand Challenge in 2016 and was also the recipient of the first ever DEF CON Black Badge awarded to a non-human entity - pioneered the application of automation, and now AI, to "offensive" security techniques. Over the years, it has developed and honed a platform that delivers continuous AI-enhanced security testing across application programming interfaces (APIs), code and software bills of material (SBOMs)."

"It also provides reinforcement learning environments for builders of foundational large language models (LLMs) to train AI agents to run, break and test software all on their own. Bugcrowd said that by folding Mayhem's platform into its own operation it will augment the ingenuity of its small army of freelance ethical hackers with the speed and precision of AI-powered testing."

""By integrating Mayhem's capabilities into the Bugcrowd Platform, we're building the industry's first truly adaptive security platform, enabling customers to anticipate, test and defend at unprecedented scale," he said. "This is a strategic step toward realising our vision of a self-learning platform that unites human creativity with machine intelligence, while shrinking customers' attack surface.""

Bugcrowd acquired Mayhem Security, an AI and cyber scaleup founded at Carnegie Mellon in 2012 as ForAllSecure. Mayhem won the DARPA Cyber Grand Challenge in 2016 and received the first DEF CON Black Badge awarded to a non-human entity. Mayhem pioneered applying automation and AI to offensive security techniques and developed a platform delivering continuous AI-enhanced security testing across APIs, code, and software bills of material (SBOMs). Mayhem also provides reinforcement learning environments for builders of foundational LLMs to train AI agents to autonomously run, break, and test software. Integrating Mayhem's platform into Bugcrowd aims to combine human creativity with machine intelligence to build an adaptive, self-learning security platform that anticipates, tests, defends at scale, and shrinks customer attack surfaces.

Read at ComputerWeekly.com