AI-based chatbots have become a transformative force driving productivity and innovation across various industries and use cases. We are at the forefront of this revolution with our AI chatbots. By continuously integrating advances in large language models (LLMs) into our platform, we are also constantly improving our customers’ user experience.
However, in highly regulated sectors such as financial services, the use of AI chatbots is hampered by considerable data protection and security concerns. Adhering to European regulations such as GDPR and DORA as well as strict compliance standards is a challenge.
Confidential computing: a new paradigm for AI security
This is where Confidential Computing comes in, offering a ground-breaking solution to maintain an unprecedented level of data protection and security.Confidential Computing is a new technology based on the latest CPUs from Intel and AMD and also keeps the data encrypted during runtime. In addition, the integrity of workloads is verified by cryptographic certificates in a process known as remote attestation. Recently NVIDIA introduced this technology with its H100 accelerator for graphics processors, extending the scope of confidential computing to AI applications.
Edgeless Systems, a leader in confidential computing, has developed Continuum AI, a framework for the secure deployment of AI models that utilizes the aforementioned H100 GPUs. Here you can find NVIDIA’s blog post about the collaboration. Continuum AI integrates confidential computing and advanced sandboxing technology to ensure that all data, including user requests and responses, remain encrypted and shielded from model owners and infrastructure providers. You can here try chatting with encrypted prompts and responses in a public preview. Continuum can also be easily integrated into our chatbot solutions via an API.
The partnership: Kauz and Edgeless Systems
Kauz.ai is working with Edgeless Systems to leverage Continuum and enable the use of LLM-based chatbots with confidential data, where the data always remains encrypted. This collaboration aims to offer our customers in highly regulated industries maximum data security. We see this as a significant milestone for the integration of AI and data protection, paving the way for wider adoption of intelligent chatbots in regulated industries. Customers who would like to find out more can contact us.
