Google Unveils Private AI Compute, a Gemini-run Cloud Platform Promising Security

Google has announced a safer way to use its AI with the Private AI Compute, offering a Gemini-run cloud platform that promises the same security as on-device processing for machine learning needs.

Google Unveils Private AI Compute Run by Gemini

Google revealed in its latest blog post that it is now prioritizing safety in the use of its AI cloud systems with the new Private AI Compute, its latest AI processing platform featuring cloud-based Gemini models.

According to Google, this new Private AI Compute features the "same security and privacy assurances you expect from on-device processing" as part of its ongoing commitment to AI safety.

The internet company claims that this new cloud-based processing platform centers on delivering on-device processing to AI use via the cloud but goes beyond its current capabilities.

According to Google, users may unlock the "full speed and power of Gemini cloud models" but at the same time, ensure that their personal data stays private and would not be accessible to anyone else, including the company.

The new cloud processing platform from Google sounds similar to what Apple offered on day one of its release of Apple Intelligence.

According to Google, the Private AI Compute was designed from the ground up, centering on core security and privacy principles which include running it on "one integrated Google tech stack" and making it private to users that no one else could access via a "hardware-secured sealed cloud environment."

Private Cloud Processing for AI

Many people have feared how artificial intelligence has significant access to a person's data, which gives it access to a person's digital profile and grows further the more they use it.

While many companies promise safety and security with the use of their AI, Apple was the first to put a great focus into a dedicated cloud processing platform, which was part of their strategy since day one.

This feature has been replicated by other companies with their unique approach to AI cloud processing, with Meta previously unveiling the Private Processing system earlier this year for all things Meta AI on WhatsApp.

However, not many companies have caught up with Apple in this privacy-first approach in deploying their AI models into the cloud.

Now, Google is on its way to delivering a platform-first approach to safety and security for its Gemini models to broadly expand privacy while using of its AI experiences.