We are excite to announce the launch of AutoGuard™, the first intelligent firewall to help enterprises deploy generative AI models safely and protect both users and enterprises from potential harms. The capabilities of generative AI are remarkable, but its shortcomings bring significant risks for enterprises. From eliminating bias and reducing hallucinations to enhancing safety and eliminating privacy issues, AutoGuard offers a strong layer of protection enabling safe and responsible enterprise-grade generative AI solutions. This firewall layer can be deployed between the client application and the LLM.
AutoGuard is available both as a hosted solution as well as an on-premise / private solution.
AutoGuard joins AutoTune™ as the second product in its AutoAlign™ generative AI platform. Both use the same AutoAlign framework incorporating auto-feedback fine-tuning and come with a library of off-the-shelf controls, like privacy protection, protection against confidential information leakage, gender assumption, jailbreaking protection and racial bias detection, or can be tailored with specific alignment controls. This means customers can enforce how their AI model behaves, expressed in natural language narratives. AutoGuard and AutoTune have been designed to work together or separately.
A large number of controls are available out-of-the-box:
AutoGuard can also be customized. A custom guardrail can be created as an AutoAlign control, using automated feedback based on how the customer has asked the AI model to behave. Throughout the process, Armilla’s own AI aligns AutoGuard through a series of steps: understanding the goals for the target model, generating data to capture the bounds of the model’s behavior, iteratively testing, discovering weak spots, and finally hardening its own guardrail model.
The result is a user-friendly platform that makes generative AI safer, trustworthy, and ethical while deployed. AutoGuard is currently being used by a select group of clients, including:
HR software companies applying generative AI to their HR processes but requiring fairness to be built into their solutions
Financial institutions looking to develop responsible generative AI solutions
Consulting firms dealing with confidential and sensitive data