Google has announced that it has created a new one privacy tool freely available to all.
Announcement of Magritte’s tool in Post (opens in a new tab) on the Google Developers blog, the company wrote that the launch will be the latest addition to Google Protected Computing Initiative (opens in a new tab)which the company says fundamentally changes “how, when and where data is processed to technically keep it private and secure.”
The new tool, which will be available on the open-source project’s Github repository, uses “high-accuracy” machine learning to detect identifying objects like license plates and tattoos and automatically blurs them out.
Google says Magritte is best placed to help videographers and journalists secure the privacy of others in the world around them. It was emphasized that it requires little computational effort and, thanks to its high accuracy, is a reliable time-saving tool.
Elsewhere, the less catchy “Fully Homomorphic Encryption (FHE) Transplier”, which aims to allow data scientists to compute encrypted data without being able to access personal data, has been first released last year (opens in a new tab)received new circuit optimizations to expand use cases, providing lower computational cost.
These tools are the latest examples of Google’s focus on privacy-enhancing technology (PET) research and development, which will launch in June 2022. has become a fire (opens in a new tab) for the Office of Science and Technology Policy (OSTP) of the United States government.
In November 2022, the concept was also presented as part of the project Competition (opens in a new tab) run by the US and UK governments, which asked participants to develop solutions to train AI models without disclosing personal information, a principle known as differential privacy (opens in a new tab).
In 2019, Google created its Differential Privacy Library – a set of tools designed for ease of use by developers and in some cases, such as Beam privacy (opens in a new tab) privacy framework, absolute non-experts – available on GitHub (opens in a new tab).