Google Cloud today introduced Policy Intelligence in beta, a series of AI-powered features designed to suggest actions by administrators aimed at improving security for organizations.

Policy Intelligence, Google’s initiative to use machine learning to enhance data governance and cloud policy stipulations for an organization, will initially be available in alpha in Google Cloud’s Identity and Asset Management (IAM). Policy Intelligence will launch with three services:

  • IAM Recommender will use machine learning to understand the different ways users typically access Google Cloud Platform and recommend roles for end users, erring on the side of least privilege to limit access.
  • Troubleshooters enables security administrators to understand why an access request was denied in an attempt to save them the trouble of wading through access policies that apply for an individual.
  • Validator helps admins set up compliance and security guardrails to prevent granting overly permissive access.

Policy Intelligence represents a new way to ingest the data of an environment to improve system performance, so administrators don’t have to rethink best practices for specific roles, Google Cloud engineering director Jennifer Lin told VentureBeat ahead of the news, describing how Policy Intelligence could continue to bring ML to governance.

“I think that’s where we can start to put some more of our best practices,” she said. “I think that whole notion that essentially as you get more and as you evolve more as an organization, you don’t have a static config or policy. Essentially it has some context awareness, and you can essentially have some situational awareness for your platform itself.”

Policy Intelligence is one of a series of security-related news items shared today at the second day of the Google Cloud Next developer conference in San Francisco.

The AI Platform, a collaborative environment for data scientists and engineers to test, build, and deploy AI, made its debut today, alongside AutoML Video and AutoML Tables for people to automatically create AI models using structured data without the ability to code.

Also announced: Android phones can now be used as security keys for Google, Cloud Platform, and G Suite accounts.