Crowds have always been a powerful thing, but before the Internet came along, it was difficult to harness them. Now things have changed: almost anything can be powered by crowds these days, from funding initiatives to news coverage, research and more. But is crowd-sourcing the right approach to penetration tests? Some people think so.
According to a report by Bugcrowd, there are literally thousands of crowd-sourced security programs today, attracting clients that range in size from small businesses to publicly traded enterprises like Motorola Mobility. And while these programs offer a number of services, the most popular one is “penetration testing” – or at least, something which goes by that name.
The fact is, crowd-sourced penetration testing isn’t like the non-crowd-sourced version at all. And while there are advantages to each approach, there are also good reasons to choose the latter over the former. To understand why, we have to start by explaining the differences between them.
Crowd-sourced vs. Traditional Pentesting
The goal of a penetration test (or pentest) is to find, document and score vulnerabilities in an information system before they are used by hackers or other malicious agents to gain unauthorized access. To do this, a pentester approaches a system just like a hacker would, from conducting reconnaissance to attempting simulated “attacks” that confirm whether a detected vulnerability is really exploitable.
Traditionally, an organization defines the goal of a pentest and hires a team of security professionals to conduct it over a limited period of time. During a crowd-sourced pentest, an organization offers a bounty to anyone who can discover a vulnerability on their systems, often through an agency with access to thousands of white-hat hackers who may or may not be professionals.
There are some advantages to the crowd-sourced model:
- Timeframe – crowd-sourced pentests take place over an undefined timeframe and may carry on indefinitely. This allows new vulnerabilities to be discovered as an organization continues to develop and improve its systems.
- Cost – a crowd-sourced pentest istypically cheaper than the traditional kind, since organizations are paying for each discovered vulnerability rather than for the test itself.
In many ways, crowd-sourced pentesting is similar to the bug bounty programs that companies have used for years to find flaws in their online platforms – and, in fact, many startups in the security industry started out as bug bounty agencies. But what works well in one context may not work well in another, and that brings us to the problems in the crowd-sourced model.
The Dangers of Crowd-sourced Pentesting
Crowd-sourced pentesting – no matter how it’s advertised – is the organized practice of inviting real hackers to hack your company and helping them to get started. Because websites are public-facing assets, offering a bug bounty does not expose them to any vulnerabilities they didn’t face before. Meanwhile, crowd-sourced pentesting requires organizations to actually connect internal systems with public channels, potentially exposing sensitive data and intellectual property to a group of individuals who suffer from:
- A lack of ethical obligations – traditional pentesters are held to a high ethical standard because their careers depend on it. They cannot hide from suspicion or blame when something goes wrong. Meanwhile, crowd-sourced hackers are often anonymous to their clients, and – while they may be required to sign a contract – in practice nothing can stop them from hiding their discoveries, or using what they find in a malicious way.
- A lack of professionalism – since crowd-sourced pentesting agencies require a large volume of talent, the quality and experience of the “hackers” they contract is wildly inconsistent. Moreover, today’s hackers often work in groups, and that’s why traditional pentesters do likewise; crowd-sourced pentesters may be lone-wolves that compete with one another for profit, generating conflict when two individuals find one vulnerability at the same time.
- A lack of focus – when an organization defines a pentest engagement they typically have a clear view of what they want to address in the test and have defined rules-of-engagement. The crowd-sourced approach tends to lack that focus and the results may be very inconsistent with the organization’s objectives.
In short, crowd-sourced pentesting removes the vital element of control that organizations normally exercise over their security operations. For this reason, companies who do invest in crowd-sourced programs – including Google, Mozilla and Facebook – also retain traditional pentesters to protect their most vital internal systems, and only use crowds where the danger does not outweigh the cost savings.
Why Crowd-sourcing is Really Popular
Aside from the low cost and flexibility that it provides, crowd-sourced pentesting is gaining in popularity due to a perception that professional pentesters aren’t “real hackers”. It is an understandable assumption: as time goes by, pentesting as a field has become dominated by automation which simply cannot rise to the human capacity for creativity and disruption.
We’re not here to deconstruct the term “real hacker” or call it a meaningless construct, because it’s not.
Hackers are not predictable. Unlike security professionals in many other fields, they do not take a linear or hierarchical view of information systems. They do not work from a CVE list, manual or rule book. Therefore, hiring a company that claims to provide “real hackers” might seem like a good solution. But real hackers are also as likely to be found working as traditional pentesters as they are anywhere else.
A Better Solution
The best hackers in the world know how to use their talents to make a sustained and comfortable living. They neither spend their days running from the law, nor do they troll the web looking for quick profit or glory. The best hackers are genuinely invisible, hiding in the very places where many assume they can’t be found.
At Securicon, we take pride in our exclusive team of bright-minded hackers from commercial, DoD and federal security backgrounds. We turn down 90% of applicants, because our pentesting program is reserved for the best and brightest in the business. We only accept talents with the right mindset for this unique occupation: they can find windows of opportunity where scanners and lesser minds see a blank wall.
The Bottom Line
At best, crowd-sourced pentesting works in a limited range of scenarios. It can help to secure production systems and other addresses that are not directly linked with your organization. However, it’s far from the best way to find vulnerabilities in your vital assets: trained penetration testers are hackers who have the intelligence, experience and creativity that it takes to find problems by working together, and the ethics to report them responsibly.
Securicon’s risk management solutions are based on industry standards for safety and professionalism. With years of experience in cybersecurity, we are here to help you manage the risks for Industrial Control Systems. Contact us for more information.