MIT’s protocol to shield cloud-based server data
MIT (Massachusetts Institute of Technology) has developed a new protocol to protect data from attackers while cloud computing without affecting the accuracy of Deep Learning Models.
MIT researchers have developed a new security protocol to secure the data, which is sent to and from cloud, while deep learning computations. Researchers have influenced the quantum properties of light to ensure the security of data. This protocol makes the most of the fundamental policies of quantum mechanics. It encodes data into laser light which avoids detection of information by cyber attacker.
Why was this protocol needed?
The continuous usage of cloud computing puts a several threat to the data security, especially in the healthcare sector where it is quite risky to use the AI tools concerning the privacy of data related to patients. The cloud- based computation is mainly focused on two parties – A client with confidential data and a main server that controls a deep learning model.
That’s why the MIT researchers have developed this protocol to ensure the security of data from cloud-based servers during deep learning computations.
“Deep learning models like GPT-4 have unprecedented capabilities but require massive computational resources. Our protocol enables users to harness these powerful models without compromising the privacy of their data or the proprietary nature of the models themselves,” said Lead Author Kfir Sulimany, MIT Postdoc in the Research Laboratory for Electronics (RLE).
Deep learning models:
Deep learning is a method for teaching computers to comprehend data as the human brain can. Deep learning models are designed to decode complex images, texts, sounds, videos, etc., accurately compared to human intelligence.
Deep learning models are being used in every for intense computations. These models require usage of powerful cloud- based servers.
How does the protocol work?
In an interview with Tech briefs, Shri Krishna Vadlamani, MIT Postdoc, briefed about the mechanism of this protocol. Shri Krishna Vadlamani said, “In simple terms, our protocol uses the quantum properties of light to secure the communication between a client (who owns confidential data) and a server (that holds a confidential deep learning model).
The server encodes the deep learning model’s parameters into light waves and sends them to the client, which performs calculations on their private data using the encoded model. The quantum nature of light ensures that the incoming model cannot be copied or intercepted by the client or any eavesdropper without subsequent detection by the server.
After the client performs the computation, the light is sent back to the server to pass verification checks; this protocol ensures that both the client’s data and the server’s model remain secure. The protocol leverages the no-cloning theorem from quantum mechanics to ensure that no sensitive information is leaked during the process.”
While ensuring the security Lead Author Kfir Sulimany, MIT Postdoc in the Research Laboratory for Electronics (RLE) guaranteed a two-way security; from the client to the server and vice – versa. Further talking about the future, the researchers desired to experiment with this protocol that how it is applied to federated learning, a multiple parties use their data to train a deep learning model.