Twitter and the Fight Against Hate Speech


Twitter and the Fight Against Hate Speech

Twitter: Championing a Hate-Speech Free Platform

In the era of digital communication, social media platforms have become a global town square, where everyone has a chance to express their views. Yet

Baidu’s Ernie Bot: The Future of AI-Powered Search Engines in China
Meta Launches AI Model LLaMA to Help Researchers Improve Chatbots
Elon Musk Reclaims Title of World’s Richest Person

In the era of digital communication, social media platforms have become a global town square, where everyone has a chance to express their views. Yet, with such freedom of speech, often comes the dark side of the coin – hate speech. Twitter is one platform taking significant strides to address this concern, with a primary focus on curating a healthy space for both content users and advertisers.

The Landscape of Twitter User Content

Twitter is a vibrant social media platform where billions of tweets are exchanged every day. From these billions, more than 99% of the content that users and advertisers see is deemed healthy. And the proportion of hate speech? An extremely small fraction.

Data on Hate Speech on Twitter

Understanding Twitter’s Own Estimates

From January to May 2023, Twitter estimated that English-language hate speech impressions constituted a mere 0.012% of the total daily impressions. These numbers are already encouragingly low, but they were soon to be challenged.

AI Stock to Watch: VERSES AI, Ticker VERS

The Independent Assessment by Sprinklr

Twitter’s commitment to creating a brand-safe environment led to a collaboration with Sprinklr in March 2023. This partnership’s goal was to leverage Sprinklr’s independent AI-based model to better understand, measure, and ultimately, reduce the prevalence of hate speech on Twitter.

The Partnership Between Twitter and Sprinklr

The Role of Sprinklr

Sprinklr, a customer experience management platform, brought their AI expertise to the table. They independently assessed the scale of hate speech on Twitter and compared their findings to Twitter’s own model.

AI in Reducing Hate Speech

Artificial Intelligence has shown promise in moderating content and identifying potential hate speech instances. By incorporating this technology, the aim was to reduce harmful content and create a healthier Twitter environment.

Progress and Results from the Collaboration

The Impact of the Partnership

Sprinklr’s AI-based model determined that the reach of daily English-language hate speech impressions on Twitter was even lower than Twitter’s initial estimates. The average daily number was found to be just 0.003% during the period of January 1, 2023 to May 31, 2023.

Pre-acquisition vs. Post-acquisition Statistics

Furthermore, hate speech impressions were noted to be 30% lower on average compared to before the acquisition, demonstrating the positive impact of this collaboration.

The Freedom of Speech, Not Reach Policy

Twitter also expanded their “Freedom of Speech, Not Reach” policy enforcement in June. This policy is designed to dramatically reduce impressions on harmful content, contributing to a decrease in hate speech.

The Future of Twitter’s Content Moderation

The Promise of Continuous Improvement

While these numbers are already impressive, Twitter’s work in reducing hate speech doesn’t stop here. The social media giant is committed to maintaining the momentum and continuing the progress towards a hate speech-free platform.

The Goal of a Healthy Global Town Square

Twitter’s ultimate goal? To become a global town square that is open for everyone, while also being safe, welcoming, and inclusive. They envision a space where everyone can truly be themselves without fear of hate or discrimination.


Twitter’s partnership with Sprinklr represents an impressive stride in reducing hate speech and fostering a healthier environment for content users and advertisers alike. The collaboration demonstrates the power of AI in creating safer digital spaces and sends a powerful message to the world – that it is possible to maintain freedom of speech while curtailing the reach of hate speech.


  • comment-avatar
    William 12 months ago

    While Twitter’s attempts are commendable I don’t think those very low numbers accurately depict reality on the platform. I’m curious if and how will AI work to combat this and not cause issues to accounts that are not spreading hate.

  • comment-avatar

    Twitter is not filled with hate but it does have accounts that continually spread lies, false information and dividing opinions. And all of this leads to hate.

    • comment-avatar

      Isn’t it nice to see Twitter patting itself on the back like this? I think they are missing the point. Just like you said, there are many accounts that post all sorts of dangerous, unproven affirmations on a daily basis and those don’t get penalized.