Skip to main content
Add as a preferred source on Google


Google Debuts Troll-Fighting AI Tool to Moderate Online Comments

Image Credit: Google Jigsaw.

Trolls are commonplace online, with close to a third of American internet users witnessing harassment online and nearly half personally experiencing it. To help combat internet trolls, Google’s technology incubator Jigsaw and Counter Abuse Technology Team last Friday launched Perspective, “an early-stage technology that uses machine learning to help identify toxic comments,” according to the blog post announcement.

Perspective works by reviewing comments and scoring them based on how similar they are to comments that are typically considered negative or “toxic,” the blog post explained. “To learn how to spot potentially toxic language, Perspective examined hundreds of thousands of comments that had been labeled by human reviewers. Each time Perspective finds new examples of potentially toxic comments, or is provided with corrections from users, it can get better at scoring future comments.”

Available through an API, Perspective is intended for members of the Digital News Initiative and other publishers and platforms. The New York Times, The Economist and more news organizations have already tested Perspective for their comment sections and found that it has significantly reduced labor-intensive work for staff, who have to sort through tens of thousands of comments each day.

Ultimately, however, publishers can decide what to do with the API, Google’s Conversation AI product manager told Wired. The tool could easily be adopted for higher education and K–12 district websites, social media pages and anywhere else comment sections exist, as a way to prevent online harassment and cyberbullying. Google said in the post that it hopes the tool will help improve conversations online.

Perspective is available in the TensorFlow open source library and Google’s Cloud Machine Learning Platform. It is in the early stages of research and development and interested developers can request API access.

To learn more, visit the Google blog.

About the Author

Sri Ravipati is Web producer for THE Journal and Campus Technology. She can be reached at [email protected].

Featured

  • programming code and digital gears

    NVIDIA Intros Open Source Tools for Building and Deploying AI Agents

    At its recent GTC 2026 conference, NVIDIA rolled out a new open source software package designed to help organizations build, deploy, and manage AI agents.

  • cyber security padlock

    AI Adoption Forces Trade-Off Between Speed and Identity Security, Study Finds

    AI adoption is forcing enterprises to trade security for speed — and identity controls are the first casualty, according to a new report from Delinea, a provider of identity security solutions for both human and AI agent identities.

  • large group of college students sitting on an academic quad

    Student Readiness: Learning to Learn

    Melissa Loble, Instructure's chief academic officer, recommends a focus on 'readiness' as a broader concept as we try to understand how to build meaningful education experiences that can form a bridge from the university to the workplace. Here, we ask Loble what readiness is and how to offer students the ability to 'learn to learn'.

  • Digital Network of User Profiles and Data Connections

    Microsoft, RSA Make Identity Security Push in the Age of AI

    Two of the bigger authentication announcements to come out of the recent RSA Conference both point in the same direction: Organizations need a more flexible, unified approach to identity security, especially as AI agents start acting alongside human workers.