Social media website Pinterest has become the first company to ban “climate misinformation” from its platform in a crackdown on freedom of expression online.

The site will no longer permit posts that oppose “the existence or impacts of climate change” or allow its users to question “well-established scientific consensus”.

Pinterest confirmed it will use machine learning technology to identify and remove posts that break its new rules, raising fears of automated overreach.

Campaign group Big Brother Watch condemned the move, warning it would be “counterproductive” and encourage distrust of mainstream social media sites.

Mark Johnson of Big Brother Watch, said: “Rather than reach for the blunt tool of censorship to shut down views which are uncomfortable or unpleasant, platforms should strive towards rules that are compliant with human rights standards.”

Sarah Bromma, Pinterest’s head of policy, said the new rules would help it “combat misinformation and create a safe space online”.

Pinterest has about 430 million users worldwide. The site has taken a similar stance against what it calls public health misinformation, cracking down on users who post information questioning measures brought in during the Covid-19 pandemic.

The social media site lets its users curate pictures, typically of lifestyle and fashion items, and has become a popular destination for arts and crafts hobbyists as well as businesses catering to them.

The new ban on climate information will apply to user content as well as advertisements.

Pinterest’s mostly automated content moderation has at its heart an artificial intelligence system capable of identifying text within images and of assessing the contents of so-called “boards”, groups of user posts.

Vishwakarma Singh, a top researcher at Pinterest, told the VentureBeat news website last year that the system can block images from appearing in users’ search results as well as their homepages.

Other social media sites have stopped short of total bans on certain topics, having faced backlash from freedom of speech campaigners.

Facebook, for example, allows users to post on controversial topics but places labels and links around such posts encouraging readers to view site-approved information.

Similarly, Twitter appends text saying “misleading” or “stay informed” to objectionable content but in most cases allows their posting.