Both ChatGPT and Google Gemini demonstrate significant potential for performing policy checks within CI pipelines, leveraging their advanced natural language understanding to interpret rules and code. ChatGPT, with its extensive general knowledge base, can effectively identify common security vulnerabilities or compliance issues based on well-defined policy descriptions. Google Gemini, especially its more code-centric or multimodal versions, may offer an edge in understanding complex code constructs and their adherence to specific coding standards or architectural policies. However, the performance of both largely depends on careful prompt engineering and the clarity of the policy definitions, as they can still struggle with nuanced interpretations leading to false positives or negatives. For reliable CI integration, robust API access and the ability to provide contextual explanations for policy violations are crucial features. Ultimately, the optimal choice often hinges on the specific policy domain, required level of code comprehension, and the effort invested in customizing models with organizational guidelines. More details: https://www.kieroads.cz/ads/www/delivery/ck.php?oaparams=2__bannerid=45__zoneid=12__cb=00b7c01792__oadest=https://4mama.com.ua