Toxic content on Twitter lower than in-house estimates: Report

Twitter shared an analysis which concluded that the reach of toxic content on the platform was lower than its own estimates

File photo of the Twitter logo | Photo Credit: REUTERS

The official Twitter Safety account on Wednesday shared the results of an independent assessment which concluded that the reach of toxic content on the platform was lower than Twitter’s own first-party estimates.

The Twitter Safety account said it had partnered with the unified customer experience management platform Sprinklr, which used an AI-powered model in order to analyse every English tweet publicly published between January and February 2023 while armed with a list of 300 English-language slur words.

What's driving the difference? The context of conversation and how we determine toxicity.

Sprinklr defines hate speech more narrowly by evaluating slurs in the nuanced context of their use. Twitter has, to this point, taken a broader view of the potential toxicity of slur usage.

— Twitter Safety (@TwitterSafety) March 21, 2023

Around 5,50,000 tweets containing at least one word from the list were found. The analysis did not report on findings related to hateful images or symbols. Twitter also pointed out that Sprinklr’s definition of hate speech was narrower than its own.

“Our focal metric is hate speech impressions, not the number of Tweets containing slurs. Most slur usage is not hate speech, but when it is, we work to reduce its reach. Sprinklr’s analysis found that hate speech receives 67% fewer impressions per Tweet than non-toxic slur Tweets,” said the Twitter Safety account.

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

The news comes as The Washington Post reported the results of a separate study on Monday which claimed that antisemitic tweets more than doubled on Twitter in the months following CEO Elon Musk’s takeover.