Just how effective is YouTube's new fact-checking feature?

Credit: Pixabay

YouTube has been working to combat the spread of misinformation on its platform by trialling a fact-checking feature in India.

Users who are part of the pilot see information panels appear in search results when they are looking up topics typically subject to misinformation. Publishers, including the Times of India, AFP and India Today, have helped contribute to the system by providing content to debunk misleading claims.

The feature first launched in the country in March 2019, with YouTube intending to extend it to other countries throughout the rest of the year.

The need for a feature like this has been clear for some time, with a majority of results for terms such as ‘geoengineering’ and ‘chemtrails’ spreading conspiracy theories and views opposing the scientific consensus, according to a study published in Frontiers in Communication.

However, Claire Wardle, executive director at First Draft, says that a lack of independent research looking at data from the initiative means there are many unanswered questions about how effective it has been.

"I have heard that people in India think it’s a positive step, but without independent review, we can’t tell what happens when people see these ‘fact-check’ labels.

"As it’s a pilot, it’s really important that we do have that research, so we can understand its impact before it gets rolled out further."

She added that the labels, similar to Facebook’s third-party fact-checking initiative, do not address issues around authentication of videos posted, with no ability to verify the time and location of where a video was filmed.

"Video content is often not about claims or 'facts', it’s about investigating whether a video has been manipulated."

Labels debunking misinformation also do not appear on videos themselves, leaving open the prospect of users stumbling across videos in their recommendations in their feed. 

However, YouTube has worked to tackle this by altering its algorithm in certain markets to reduce the presence of conspiracy theories. This update currently being rolled out to users in the UK.

Since changes to the recommendation system came into effect in the United States in January, views of so-called 'borderline content', videos that misinform but do not violate the website’s guidelines, have dropped by more than 50 per cent.

But even trying to define 'misinformation' is a challenge when determining what content falls into this category, Wardle explained.

"Are you talking about conspiracy, outright lies, misleading content, or outright false content?"

Rather than having tech giants developing their own approach to tackling the spread of misinformation, Wardle would like to see a centralised solution.

"It makes no sense for each of these platforms to have their own programs, and until we know whether it works, we can’t assess how effective any sort of ‘fact-checking’ initiative is."

Save the date: Newsrewired takes place on the 27 November at Reuters, London. Head to newsrewired.com for the full agenda and tickets

Free daily newsletter

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).