Kong has introduced updates to its AI Gateway, a platform for governance and safety of LLMs and different AI assets.
One of many new options in AI Gateway 3.10 is a RAG Injector to cut back LLM hallucinations by routinely querying the vector database and inserting related knowledge to make sure the LLM is augmenting the outcomes with recognized information sources, the corporate defined.
This improves safety as properly by placing the vector database behind the Kong AI Gateway, and likewise improves developer productiveness by permitting them to give attention to issues aside from trying to cut back hallucinations.
One other replace in AI Gateway 3.10 is an computerized personally identifiable info (PII) sanitization plugin to guard over 20 classes of PII throughout 12 completely different languages. It really works with most main AI suppliers, and might run on the world platform degree in order that builders don’t must manually code the sanitization into each utility they construct.
In response to Kong, different related sanitization choices are sometimes restricted to changing delicate knowledge with a token or eradicating it solely, however this plugin optionally reinserts the sanitized knowledge into the response earlier than it reaches the top person, making certain they’re able to get the info they want with out compromising privateness.
“As synthetic intelligence continues to evolve, organizations should undertake strong AI infrastructure to harness its full potential,” mentioned Marco Palladino, CTO and co-founder of Kong. “With this newest model of AI Gateway, we’re equipping our clients with the instruments essential to implement Agentic AI securely and successfully, making certain seamless integration with out compromising person expertise. Furthermore, we’re serving to clear up among the greatest challenges with LLMs, reminiscent of chopping down on hallucinations and bettering knowledge safety and governance.”