Facebook is Thinking to Only Demote COVID-19 Misinformation Posts Instead of Removing Them

Social media network Facebook may stop removing COVID-19 misinformation on the platform and demote them instead.

Meta, Facebook's parent company, announced that it is considering the move as it seeks advice from the independent oversight board on whether to modify its COVID-19 misinformation policy.

Facebook's COVID-19 Post

According to PCMag, Facebook expanded its policy on harmful COVID-19 misinformation in 2020 as the virus spread around the world, allowing for posts that could lead to an "imminent risk of physical harm" to be removed worldwide instead of only being taken down when experts point it out.

The policy was designed by the platform to combat misinformation about the COVID-19 pandemic, like false claims about the effectiveness of social distancing, masks, and transmissibility of the virus.

In late 2020, as the first doses of the COVID-19 vaccines started rolling out, Facebook updated its policy to remove any misleading or fake information about the vaccine, according to CNET.

According to Meta, they removed more than 25 million posts since the pandemic.

But now, amid the shifting trends and declining stocks of Meta in 2022, the company is looking into changing its policy, starting with input from its oversight board.

The company said that its Community Standards policies protect free expression while preventing dangerous content from being seen in its platform.

Meta admitted that resolving the tension between free expression and safety is not easy, especially when confronted with unprecedented and fast-moving challenges, as what they've seen during the pandemic.

Facebook's content moderation guidelines have been a contentious topic, with the company being accused of enabling hate speech and misinformation on the site to optimize profit and limiting free speech.

Meta's independent oversight board comprises legal advisers from several think tanks, professors from universities worldwide, journalists, and human rights advocates.

In this case, the company requested an advisory opinion from the oversight board, meaning the recommendations they give to Meta are nonbinding.

Meta's AI Tool

As Facebook announced its plan to stop removing COVID-19 misinformation on the social media site, its parent company revealed that it has a new AI tool that detects misleading posts.

The new AI-powered tool, Sphere, is intended to help detect and address misinformation on the internet.

Meta claims that it is the first AI model that is capable of automatically scanning thousands of citations at once in order to check whether they support the claims or not.

According to CNBC, Sphere's dataset includes 134 million public web pages. It relies on that collective knowledge of the internet to rapidly scan thousands of citations in search of factual errors.

Meta is training Sphere using entries on Wikipedia. The company said that the AI tool is already scanning pages on the crowd-sourced internet encyclopedia to test its ability to flag sources that do not support the claims in the entry.

The social media company also said that when the AI tool spots a questionable source, it can recommend a stronger one to help improve the entry's accuracy.

This article is owned by Tech Times

Written by Sophie Webster