As application workloads and activities move to one or more cloud environments, organisations have also begun to ask what they should do to ensure proper cloud security.
The answer, it turns out, has much to do with going back to basics as it does with anticipating what’s coming next in cloud-based security.
What NOT to Do
When we advise clients to “go back to basics” for cloud security, we often look first at what they are doing now to secure their current IT environments. Such practices—whether good or bad—often extend to an organisation’s use of hybrid and public cloud services as well.
For starters, we recommend abolishing some common, but risky security habits:
- Stop the use of manual, static or stale credentials. In the age of software-defined infrastructures and dynamic cloud services, it’s time to move from manual and often-unchanging user access methods. This includes:
- Stopping the use of static access keys that never expire.
- Stopping the use of access keys embedded in software code during application development or testing. These pose a risk if they remain in production code or in the event the code is subsequently uploaded to an open source repository, such as GitHub.
- Stopping the assignment of permissions that you forget to revoke when the need for access ends. Instead of assigning permissions that may last months or years, adopt a security posture that assigns permissions that last minutes or hours, especially for more sensitive and impactful data.
- Stop storing User IDs and passwords for longer than necessary. Given the growing focus on user privacy and regulations like the GDPR, it’s past time to start following some of the tenets adopted by the banking and payment card industry (PCI).
Just as those organisations don’t hold credit card information for longer than necessary, so too should today’s organisations keep user authentication data for only as long as needed, then discard it. Better yet, organisations should not try to take on the responsibility to hold user secrets they can’t protect properly. It may be time to pass on that responsibility to a third-party or cloud service expert in managing user identities.
Cloud service providers who deal heavily in user security and authentication are also increasingly moving toward a future that authenticates a user’s access via some type of trusted user account instead of via disparate, application-specific user IDs and passwords. This might include verifying user identities through a Microsoft Office 365 account, for instance, or through one of the user’s other social media networks or accounts. This means no longer creating a new User ID and password unique to your application or system. Instead, reuse trusted sources for identity and access management.
- Stop giving administrators blanket access to everything, all the time. This practice may make an administrator’s life easier, but it has left a lot of damage and breached systems in its wake. Hackers often use such security loopholes with blanket Admin credentials to gain access to sensitive corporate data. Look into new solutions that easily control this type of access without causing a major headache for the administrators. Many do so while still ensuring appropriate governance.
Good Steps to Prepare for a Secure (Yet Cloudy) Future
We’ve covered many ways organisations tend to fall short in their security efforts. But, what can you start doing now to secure your emerging hybrid and multi-cloud worlds? To start, look to what leading cloud security providers have begun doing:
- Start replacing blanket permissions, user IDs and passwords with multifactor authentication (MFA) and privileged access management (PAM).
- Incorporate condition-based access to everything for everyone. The future will allow normal users to go about their day-to-day work without being constantly prompted for credentials. This will occur by monitoring behaviours and making appropriate decisions about when to escalate authentication and prompt for more layers of verification.
- Look to tools that use AI and machine learning to help evaluate log data and help flag the difference between typical user behaviour and questionable behaviour that might fall outside the norm.
- Start incorporating single sign-on for your users and customers, preferably from one cloud-based provider. For example, if you are using something like Active Directory in your on-premises environment, look for a provider that lets you automatically synchronise user login IDs/passwords from that system to a cloud counterpart. Similarly, consider offloading the hosting of user IDs and password databases to a reputable cloud provider.
- Consider how to securely collaborate with business partners, third-party vendors and contractors vs. creating their own user accounts on your network.
If you are confused about what to do first, prioritise efforts based on low-hanging fruit. This means protecting and securing the highest risk users and data in your organisation. This could be your executives, your HR team, your Finance team, or other IT administrators. Focusing in this way will reduce your risk exposure significantly. It will also teach you about how to apply these controls to the rest of your organisation.