After Elon Musk’s takeover of Twitter, content moderation to prevent hate speech is back in the spotlight. While Twitter could be slipping into moderation chaos, Mark Zuckerberg’s Meta says dealing with hate speech on Facebook and Instagram is on the right track.
Pexels
Eliminate hate speech on Facebook, Instagram
Meta says hate speech detection has dropped dramatically on Facebook and Instagram. In its latest Integrity and Transparency Reports, Meta claims that its hate speech detection rate increased from 95.6% to 90.2% in the third quarter of 20233.
We publish our latest reports on how we apply our policies in @Facebook and @instagram. We have continued to progress in:
👉Keep the prevalence of infringing content low
👉Protect people from deceptive networks of fake accountshttps://t.co/sEWtMKnt90— Meta Newsroom (@MetaNewsroom) November 22, 2022
That’s not all! The company says that for content related to bullying and harassment, Meta saw a drop from 76.7% to 67.8% on Facebook and from 87.4% to 84.3% on Instagram.
Read also : Meta’s New AI System “Galactica” Can Write Research Papers From Scratch
The report also adds that Meta removed 16.7 million content (some of which was misreported and later restored) related to terrorism, 4.1 million drug posts and 1.4 billion spam content on its platforms.
Reuters
Meta’s AI can now better detect funny terms that can be used between friends. Additionally, it can detect words that may be considered offensive/inappropriate in one context but not in another. “As we improve this accuracy, our proactive hate speech detection rate has also decreased from 95.6% to 90.2% in Q3 2022.”
It looks like Meta has also improved ways to identify and block spam on Facebook, including removing users who engage in “inauthentic behavior.”
Pexels
Last year, Meta’s Facebook was in the vortex of hate speech and misinformation. Now the company seems to have made progress on this front.
Read also : Shadow to Meta? Apple Vice President Greg Joswiak Says He Will Never Use the Word “Metaverse”
Meanwhile, companies like Twitter don’t seem to be faring too well on this front, especially after Elon Musk let go of a sizable chunk of staff and introduced features prematurely (like paid verification), adding to the narrative. of hate and the platform’s impersonation problem. .
What do you think of the slowdown in hate speech on Meta platforms? Let us know in the comments below. For more in the world of Technology and Sciencekeep reading Indiatimes.com.
References
Gairola, A. (2022, November 23). Meta says hate speech content has decreased on Facebook and Instagram – is Elon Musk’s Twitter listening? -M. Benzinga. https://www.benzinga.com/news/22/11/29834512/meta-says-hate-speech-content-decreased-on-facebook-and-instagram-is-elon-musks-twitter-listening
The Hindu. (2022, November 23). Improved AI helps Facebook and Insta fight hate speech, violence and bullying. https://www.thehindu.com/sci-tech/technology/improved-ai-helps-facebook-meta-insta-curb-hate-speech-violence-bullying/article66173191.ece
Comments are closed.