Pentesting is a technical security check at one point in time, either with full system knowledge (white-box) or starting blind (black-box). Red teaming is different: you're given a specific goal like stealing a database or getting domain admin access, then operate within those constraints to achieve it, mimicking how real attackers work.
I would consider pen test or penetration testing as a point-in-time evaluation of an organization's technical security posture. You can test at multiple different layers and from multiple different perspectives. You can test white box, which is you have full knowledge of the environment and you know what you're getting into and you're able to conduct a very thorough test because you know what's there. You can do a black box test, which would be blind like a true adversary coming in and you just start with open source intelligence and reconnaissance, and you basically know nothing but what you are able to glean from the actual testing. Red teaming, I would consider as something much more objectives-driven. So with the red team, you are given a target. It may be domain admin, or it may be the CEO's personal bank account number, or it may be the molecule database at the pharma company. Right. And you're given a set of tactics that are allowed and then broad remit within those tactics to try to achieve that objective. Cool.
