Automated contract review

The review of data processing agreements (DPAs) is relatively standardized. However, despite the high degree of similarity in the GPAs and standardized review processes, the repetitive activity costs lawyers and legal assistants a lot of time.

The challenge in automation, however, is that the criteria to be checked can often be not only explicit but also implicit. Classical NER-based systems can therefore only be used to a limited extent and may, for example, produce incorrect results due to negations and subjunctives.

The knowledge about the correct assessment of the DPAs thus lies with the auditors and is difficult to formulate explicitly in a rule-based manner. Therefore, an AI-based system had to be developed that captures and learns the assessment and the associated rationale in the course of the audit and can be applied to new formulations.