In recent decades, various countries have enacted laws for protecting the privacy of individuals, one such example being the European General Data Protection Regulation (GDPR). The notion of purpose plays a crucial role in these regulations. However, in most situations purposes are expressed as vague textual descriptions, which often prevents (i) the user from fully understanding how their data is going to be processed, and (ii) a precise analysis that a system satisfies a given (GDPR-based) privacy policy. To address this shortcoming, various works have been concerned with providing semantic foundations to the notion of purpose. In this work, we pursue the same goal and we advocate the use of behavioral types as a means of capturing the notion of a purpose and, consequently, the use of type checking to formally verify that a system conforms to its privacy policy. We instantiate this proposal in the context of our previous work on the Privacy calculus.