The Spanish company’s technology aims to improve the accuracy of models and algorithms without sharing private data.
Sherpa.ai today announced the launch of its artificial intelligence (AI) platform applied to data privacy, based on the federated learning paradigm and allowing AI models to be trained without sharing private data.
The Spanish company stresses that the platform will allow companies and institutions to harness the potential of their data and AI, as well as explore new scenarios for collaboration between organisations in sectors such as health, finance and security, among others.
The platform has been tested and is being used by organisations such as KPMG, Telefónica and the NIH (United States National Institutes of Health), with which Sherpa.ai has already announced agreements. For example, the NIH believes that this technology will allow it to improve the algorithm for diagnosing diseases without the need to share private patient data. ‘It could make it possible to explore new diagnostic and treatment possibilities for a group of diseases that currently have no specific treatment,’ says Carsten Bönnemann of the NIH.
Sherpa.ai’s technology is based on the principles of privacy, security, and data ethics, the company said in a statement. The platform makes it possible to improve the accuracy of model predictions and algorithms, but in a private and ethical way, as the data is never shared. ‘Data privacy is a fundamental ethical value and this platform we are announcing today can represent a before and after in the use of data in a private and secure way for Artificial Intelligence developments, and has the potential to help solve major challenges such as saving lives, tackling fraud, increasing security or reducing carbon footprint,’ said Xabi Uribe-Etxebarria, founder and CEO of Sherpa.ai.
Sherpa.ai has announced the three products that will be available soon. A freemium version that will allow experiments with the platform. The Enterprise version, which will allow companies to maximise the value of their own data. And the Hub version, which will allow organisations to allow third parties to use their data without having to share it, thus guaranteeing privacy and security.
The announcement was attended by Tom Gruber, Chief AI Strategy Officer of Sherpa.ai and former CTO and founder of Siri; Thomas Kalil, former Director of Science and Technology Policy at the White House, who recently joined the team as Advisor, Carsten Bönnemann of the NIH (US Department of Health), and Ritu Jyoti, Vice President of Artificial Intelligence at IDC Group.