How to limit penetration test risks by defining testing scope

Penetration tests can obviously help harden a company's security profile, but are there any risks that stem from...

the pen tests themselves? Should companies place any limits on what a pen test can do?

There are potential risks involved with performing penetration tests, but the benefits can outweigh the risks. There are risks similar to those that can crop up during an in-depth vulnerability scan, and the exploitation steps that a skilled pen tester might take can pose additional risks. For example, using a well-known exploit could have less risk than custom-developed exploits, where the results of execution are unknown. To protect against potential penetration test risks, companies should ensure they have adequate backup plans in case an application or servers fail, or data is deleted or corrupted as a result of a pen test. Companies should also have tools in place to detect problems with the systems in the pen test so the systems can be recovered quickly.

The penetration testing scope can be limited, but some may question the results and whether the vulnerabilities identified could be exploited. Test or development systems could be used for the pen test to limit the potential risk, but these systems need to be configured like the production systems to get the most value. The tests could also be limited to off hours or during production downtime windows to limit potential risks. Carefully defining the pen testing scope can also help minimize potential collateral damage from intrusive scans or exploits. The test can be limited to identifying whether an exploit could be used to take over a system. If there are significant host-based security controls, there may be a need to try to exploit the system to determine if the other security controls in place could protect the system.