Evaluating data security and ROI
An important topic of any development lifecycle is what threats and data security controls should be considered before using AI. We have been sending things to different systems, including the Poe.com platform, which then traverses our detections and log samples to other third-party systems. Like any other system we build, it should start with a design and go through an architectural review.
CI/CD pipelines and deploying detections to security tools are typically not going to catch the attention of your security architects. However, sending things to external systems via an API might. Compensating controls that can help make the case for introducing AI-augmented testing, and even general development, include the following:
- Ensuring DLP or CASB is deployed at all developer endpoints including terminated TLS and SSH protocols for deep inspection
- Using pre-commit hooks in any development environment looking for regex or keywords that should...
