CoCoS.ai (Confidential Computing System for AI) lets you run AI/ML workloads on combined datasets from multiple organizations while guaranteeing the privacy and security of the data and the algorithms. Data is always encrypted, protected by hardware secure enclaves (Trusted Execution Environments), attested via secure remote attestation protocols, and invisible to cloud processors or any other 3rd party to which computation is offloaded.
CoCoS.ai is an innovative project funded by the Serbian government’s Innovation Fund (Collaborative Grant Scheme Program, project ID 50314). The awarded consortium is a collaboration between two member organizations: Ultraviolet, which comes from the software and cloud security industry, and ETF University of Belgrade, department of data protection and security, which is an academic partner.
Ultraviolet is a technology company formed in 2015. to offer specialized high-tech services in the domain of cybersecurity. The company is focused on confidential computing methodologies in the domains of cloud, telecom and AI/ML.
The University of Belgrade School of Electrical Engineering is one of Serbia's most important and influential academic institutions. It is the largest engineering faculty in the region and one of the largest engineering faculties in Europe.
CoCoS.ai is a distributed, microservice-based solution in the cloud that enables confidential and privacy-preserving AI/ML, i.e. execution of model training and algorithm inference on confidential data sets. Privacy-preservation is considered a “holy grail” of AI. It opens many possibilities, among which is a collaborative, trustworthy AI.Final product enables data scientists to train AI and ML models on confidential data that is never revealed, and can be used for Secure Multi-Party Computation (SMPC). AI/ML on combined data sets that come from different sources will unlock huge value.
STEP 1 - Consortium is created for confidential collaborative AI/ML computations
STEP 2 - Organizations upload their encrypted data sets into a secure enclave
STEP 3 - AI companies deploy algorithms on the combined data sets without ever seeing data (data is decrypted only in the secure enclave where the algorithm is applied)
STEP 4 - Only insight or trained model is brought back to AI company, never raw data