Coinbase conducted a detailed analysis of various ASM offerings and chose to use the IBM Security® Randori® Recon solution, including its IBM® X-Force® Attack Targeted service, which, as Lunglhofer says, “came out on top from a value perspective and from a pure capability perspective.”
Coinbase has now used Randori for several years and, as the company has rapidly evolved, it’s worked closely with the Randori team to continually test and strengthen its everchanging attack surface. Randori combines ASM automation software with X-Force Red, a human team of cybersecurity experts and ethical red team hackers. While the automation drives surface testing at scale, the people collaborate directly with Coinbase to address issues requiring deeper insight and nuance.
“I’ve always been incredibly impressed by the Randori team members that I’ve spoken with,” says Lunglhofer. “They spend a lot of time listening to our concerns, then they take that information and synthesize it into a really meaningful attack plan that shows us where we can tighten things. It really is next level, the deep understanding and spending hours on the phone talking through our business and getting very focused on the targeting. That’s a huge differentiator, that level of investment.”
Over time, Coinbase has adjusted how it uses Randori in order to maximize its effectiveness. In the first year or so, Coinbase had the Randori team operating almost independently, as a “very-red red team,” as Lunglhofer puts it, stealthily hunting across Coinbase’s network for any potential risks. There were some very valuable finds, but by and large, thanks to the skills of the Coinbase security team, the network was tight. It was difficult for the red team to find areas needing improvement.
Seeing potential for steadier value, Lunglhofer and team replaced cloak-and-dagger red teaming with more collaborative purple teaming, having the Randori red team work directly with Coinbase security “blue” teams in attack planning sessions. Now, Coinbase teammates with deep knowledge of the network guide the red team to the areas they most want to test. The result, Lunglhofer says, is “more impactful tests that more accurately simulate an adversary. It’s a much more consistent positive impact.”