FHE enables computations on encrypted data, ensuring that sensitive information is never exposed during outsourced processing.

We accelerate the practical application of FHE by:

1. Enhancing computational and memory efficiency in real-world deployment.

2. Strengthening the security and integrity of FHE-based systems.

Use Cases of FHE

Apple: Machine Learning & FHE

Microsoft: Password Monitor

MPC allows multiple parties to jointly compute functions on their private inputs without disclosing the underlying data to one another.

We enhance the real-world viability of MPC through:

1. Efficiency Improvement: Reducing communication and computational overhead to support various deployment environments.

2. Specialized protocols: Designing optimized solutions for specific tasks,
   such as Private Set Intersection and Secure Aggregation.

As ML/AI increasingly relies on sensitive data, addressing emerging privacy and security threats is paramount. 

We enhance the privacy and security of the AI ecosystem by developing advanced cryptographic solutions for:

1. Federated Learning & Secure Aggregation: Enabling distributed training and model updates without disclosing private data.

2. Secure AI-as-a-Service: Protecting sensitive user data and proprietary model information in cloud-based AI services.

3. Privacy & Security of AI: Exploring defense mechanisms for secure AI deployment against evolving adversarial threats.