Security
AI Model Security
Models Used
Kodus leverages cutting-edge language models for unstructured data processing and advanced analysis. The models currently in use are:- GPT-4o and derivatives: An advanced language model developed by OpenAI, known for its high-quality text comprehension and generation capabilities.
- Gemini 1.5 Pro and derivatives: A multimodal model from Google, capable of processing and generating text, images, and other types of data.
- Claude 3.5 Sonnet and derivatives: A language model by Anthropic, designed with a focus on safety and ethical alignment.
Data Handling
Temporary Use
All data processed by these models is used solely for temporary contextualization, which means:- Data is retained only for the duration of the processing session.
- After task completion, data is immediately discarded.
No Training Usage
In line with our agreements with model providers:- No customer data is used to train or improve the models.
- Privacy and confidentiality of information are rigorously maintained.
Terms of Use for Each Model
- GPT-4o and derivatives: GPT-4o Terms of Use
- Gemini 1.5 Pro and derivatives: Gemini 1.5 Pro Terms of Use
- Claude 3.5 Sonnet and derivatives: Claude 3.5 Sonnet Terms of Use
Security Measures
To ensure the protection of our clients’ data, we have implemented the following measures:- Encryption: All data is encrypted both in transit and at rest, using state-of-the-art encryption protocols.
- Data Isolation: Models are designed not to store or retain data after processing.
- Controlled Access: We enforce strict access controls to ensure that only authorized personnel can interact with systems processing customer data.
- Regular Audits: We conduct periodic security audits to ensure compliance with industry best practices.
- Security Updates: All systems and models are regularly updated with the latest security patches.