1sec.ai

Tag

#privacy

Every item tagged privacy, newest first.

3 items

toolsApr 27

How to build scalable web apps with OpenAI's Privacy Filter

OpenAI released a Privacy Filter to help developers build scalable web apps that protect user data. The filter enables apps to detect and redact sensitive information before sending it to OpenAI's APIs. This helps developers comply with data protection regulations like GDPR and CCPA. You can integrate the filter into your app using Hugging Face's Transformers library.

Key takeaways
  • OpenAI provides a Privacy Filter for detecting and redacting sensitive user data.
  • The filter helps apps comply with GDPR, CCPA regulations.
  • Hugging Face's Transformers library supports integration.
HHugging Face Blog#privacy#data-protection#web-apps
modelsApr 16

Running Privacy-Preserving Inferences on Hugging Face Endpoints

Hugging Face now supports fully homomorphic encryption (FHE) on its model endpoints, enabling private inference without exposing input data. This allows developers to run inferences on sensitive data while maintaining confidentiality. Builders can integrate FHE for secure model deployment.

Key takeaways
  • Hugging Face supports fully homomorphic encryption on model endpoints.
  • Enables private inference without exposing input data.
  • Developers can integrate FHE for secure deployment.
HHugging Face Blog#privacy#homomorphic-encryption#secure-inference
researchAug 2

Towards Encrypted Large Language Models with FHE

Researchers propose using Fully Homomorphic Encryption (FHE) to enable secure inference on large language models. This approach allows computations on encrypted data without decryption, preserving user privacy. Builders can integrate FHE into existing models for enhanced security. FHE-based LLMs have potential applications in sensitive domains like healthcare and finance.

Key takeaways
  • FHE enables computations on encrypted data without decryption.
  • Encrypted LLMs can preserve user privacy in sensitive domains.
  • FHE integration is feasible with existing models.
HHugging Face Blog#encrypted-llms#fhe#privacy#security