Policy Configuration Strategies

When configuring policies, consider the different types of policies that your organization requires. In addition to having both read and write policies, you might also need different flavors of read and write policies to address the different levels of security required for various endpoints or for various user groups.

For example, you might create one read policy that obfuscates most sensitive data, such as company IDs, home addresses, social security numbers, and salaries. But your Human Resources department needs access to some of that information, so you create an HR read policy that allows viewing the data.

Similarly, you might have a somewhat lenient write policy for writing to internal destination systems managed and accessed entirely by high-level organizational users. You also create more restrictive policies for systems that are accessed by lower-level organizational users, contractors, or third parties.

Before configuring policies, consider the protection methods that you want to use. Protected data can still provide value and be used for processing or reporting, depending on the protection methods that you use. Some protection methods completely obfuscate data, making it impossible to use. Other methods generalize data, providing a different form of the data that might be used.

For example, you can completely obfuscate social security numbers by replacing them with a string, "SSN redacted". But you can round birthdays to a year, protecting the most sensitive details while preserving some information for processing.