Google announced that it has made a new privacy tools It is free for all to use.
Announcing the Magritte Tool postal (opens in a new tab) On the Google Developers blog, the company wrote that the release will be Google’s Protected Computing Program (opens in a new tab)which the company claims is intended to fundamentally change “how, when and where data is processed to technically ensure its privacy and security.”
The new tool, which will be available on the open-source project repository Github, uses “high-precision” machine learning to detect recognizable objects, such as license plates and tattoos, and automatically blur them out.
Google claims Magritte is best suited to help videographers and video journalists protect the privacy of others in the world around them. It emphasizes that its computational cost is low, and its high precision makes it a reliable time-saving tool.
Elsewhere, the less glamorous “Fully Homomorphic Encryption (FHE) Transplier” aims to let data scientists compute encrypted data without access to personal information, and has been first released last year (opens in a new tab)has gained new circuit optimizations to expand its use cases by ensuring lower computational costs.
These tools are the latest example of Google’s focus on research and development of Privacy-Enhancing Technology (PET), which will be released in June 2022, become the focus (opens in a new tab) US Government Office of Science and Technology Policy (OSTP).
In November 2022, the concept is also available as Contest (opens in a new tab) Run by the US and UK governments, participants are asked to develop solutions that allow AI models to be trained without exposing personal data, a principle known as differential privacy (opens in a new tab).
In 2019, Google produced the Differential Privacy library — a set of tools designed to make it easier for developers to use, in some cases, such as Privacy on Beam (opens in a new tab) Privacy frameworks, definitely non-experts — Available on GitHub (opens in a new tab).
