Facebook in 2016 launched a set of tools designed to help those that may be at risk of hurting themselves. Developed in partnership with Lifeline, Forefront and Save.org, the suicide prevention tools were expanded globally this past spring.

Now, those tools are being automated.

Guy Rosen, VP of Product Management at Facebook, said on Monday that the social network is working on several proactive suicide detection efforts. Facebook’s AI, for example, is using pattern recognition to detect posts or live videos where someone might be expressing thoughts of suicide. The tech is also being used to prioritize the most concerning reports, bumping them up the priority list and getting local authorities involved twice as quickly.

Rosen said Facebook is also working to improve how it identifies appropriate first responders. They’re additionally dedicating more reviewers from the community operations team to review reports of suicide or self-harm.

The efforts seem to be paying off. Over the last month, for example, Facebook worked with first responders on more than 100 wellness checks based on reports received by the new proactive detection efforts.

Facebook is starting to roll the AI-based tech out to territories outside of the US. The company won’t be doing so in the EU, however, as local privacy laws restrict the use of profiling users based on sensitive information.

Lead image courtesy USO