Tech businesses are attempting to provide an effective method to prevent the dangers posed by issues like misinformation, impersonation, pornography, and violent graphic content at a time when fake news and AI-generated media content are prevalent on social media.
The Indian microblogging platform Koo is working to develop into a platform that uses machine learning and artificial intelligence (AI) to improve its content regulation.
In order to make sure that its content moderation tactics are effective and help the platform become a safe place for all stakeholders, Koo has revealed a number of features.
Although the social network's interface resembles that of Twitter, the business insists that Koo stands out because of its dedication to creating a secure and egalitarian environment. Now let us take a peep into the working methodology of the AI-powered Content Moderation tool of Koo.
How does it work on Nudity and Pornography?
The company claims that the entire procedure is automated and begins just seconds after the picture is posted. The user will receive a second message after the image has been deleted explaining why it was taken down and inviting them to submit an appeal using the redressal form if they feel that a mistake was made. These notifications will be displayed in the user's selected preferred language.
The moment someone publishes a nude image on Koo, they will receive a notification that reads, "This Koo has been deleted due to GRAPHIC, OBSCENE OR SEXUAL CONTENT."
Similar to this, if a user uploads a video that contains explicit sexual content or nudity, it will be removed in about five seconds, depending on how long the video is and how long it takes to process it. Koo will notify the user once the video has been deleted. Koo will remove a user's nude display picture if they post one, employing the same process.
What Action will Koo-AI take on Posts containing Violence?
Koo has developed a sophisticated approach and eliminated the sweeping delete mechanism previously observed in terms of obscene content because some of these photographs may be connected to a news development.
When a user uploads a graphically violent or gory image, Koo will approve the post but add an extra layer of warning. A hazy image with the words "This content may not be appropriate for all users" will then show. Their intrigue has been marked with a warning message. Users will have the option to view, like, or comment on the photograph.
How will the AI approach of Koo deal with Impersonation?
The platform uses machine learning to spot impersonation attempts, and according to the business, it is growing better at it. AI is used to support the impersonation detection process, while human moderators still handle the majority of the manual work.
The platform's impersonation dashboard, which is only intended for company employees, provides important details about the user and the VIP being impersonated. Soft Delete, one of its features, eliminates all information that could be construed as impersonation, including name and display picture.
The site notifies the user after this counterfeit content has been removed that "Your profile details are removed - Your profile details are removed due to repeat violations of the Koo community guidelines or legal requirements."
How will Koo deal with Fake News using AI?
The platform asserts that it can quickly remove fake news since it runs a detection cycle every half-hour. When a user spreads bogus news, the dashboard recognises it and provides information tracing the content's origins, providing the moderator with enough details to act right away. The phony news is accompanied by a notification that reads, in part: "Unverified or False Information: Reviews by a Fact Checker." If users think their content is authentic, they can also request a review.
How will the Koo AI algorithm deal with Toxic Comments & Spam?
The user made a derogatory comment on a post as part of the demo. Koo recognises these postings and effectively conceals them. Users must select the Hidden Comments button in order to see such postings. This function operates almost immediately. According to the business, Koo's method of allowing individuals to express their opinions is by permitting comments, even though blocking them would equate to curtailing freedom of expression.
This is another quick development of the Indian microblogging Platform, Koo, in the process of reinforcing its content moderation with AI accuracy. Very soon, it is going to turn up as one of the most secure and reliable social media site for netizens in the country!
Comments