- The free, open-source Python SDK and CLI tool complements Cado’s enterprise/commercial cloud forensics and incident response software.
- It made headlines when Cado Security raised USD 20 million in fresh capital earlier this month for global expansion and product development.
Recently, Cado Security Ltd., a cloud forensics and incident response platform startup, announced the launch of a new security solution geared towards APIs for large language models.
The open-source library known as Masked-AI provides secure use of LLM APIs like OpenAI LP’s GPT-4 without transmitting private data. It may be downloaded from the GitHub source, enabling programmers to use OpenAI APIs without worrying about security issues.
Co-founder and chief technology officer of Cado Security, Chris Doman, said, “Masked-AI is an open-source tool that anyone can use to ensure better security and privacy when leveraging OpenAI/ChatGPT, which is especially important in the context of incident response. As active participants in the security community, this is an impactful milestone for empowering security professionals to respond to threats at cloud speed. This simple yet powerful tool demonstrates that we can use LLMs securely without compromising on the quality of the responses generated.”
With the help of GPT-3, Cado Security’s platform previously provided an Interactive Incident Response tool that streamlined forensic investigations and ramped up responses. However, the business detected a number of potential issues, including privacy-related issues, with GPT-3 in production systems for the incident response use case.
Through Masked-AI, Cado Security hopes to create a community around enhancing the effectiveness of security teams while safely utilizing LLM APIs to speed up incident response in the cloud.
The service hides user names, IP addresses, credit card numbers, email addresses, and phone numbers. After replacing sensitive data with a placeholder, the tool submits the masked request to the API. To recreate the API result, the solution keeps a lookup table locally. The user can then access sensitive info after that.
The free, open-source Masked-AI Python software development kit and command line interface tool complement Cado’s enterprise and commercial cloud forensics and incident response product. The new open-source application has also been linked with the Cado platform to further streamline and speed up the incident response process.
It made headlines when Cado Security secured USD 20 million in fresh funding for international growth and product development earlier this month. Eurazeo S.A., CyLon Ventures Ltd., Ten Eleven Ventures LLC, and Blossom Capital LLP are a few of the company’s investors.