New Delhi, May 2
As global elections gather steam, the independent Oversight Board warned on Thursday that inaccurate enforcement by social media platforms and unequal investment in global election integrity efforts have increased the risks of poll-related violence, the problematic suppression of critical political speech and the spread of fake news and deepfakes that aim to interfere with elections.
The Oversight Board, an independent body of 22 global human rights and freedom of expression experts from across the political spectrum and the world, has made elections one of its main strategic priorities.
In a paper, the independent board called for urgent action to tackle the worst online abuses threatening key global elections and free speech.
The Board analysed a broad collection of its election-related cases -- ranging from the suspension of former US President Donald Trump on Meta's platforms to the Board's decision to leave up controversial comments about immigration by a French politician, as well as news reports on political speech in the Pakistani Parliament.
It then devised nine key lessons that all social media platforms should follow.
These include dedicating sufficient resources to moderating content before, during and after elections, not limiting this to the immediate voting period, setting basic standards for all elections everywhere and not neglecting dozens of elections taking place in countries that might be less lucrative markets, but where the threat of instability is often greatest.
"Our guidance has been clear: if you are taking down content relating to politics, especially in the context of elections, removal must be necessary to prevent or mitigate genuine offline harms. The right of voters to express and to hear a broad range of information and opinion is essential," said Michael McConnell, Oversight Board Co-chair.
"But this cannot be an excuse for allowing posts that will likely lead to violence or suppress the rights of others to speak and to vote," he added.
The guidelines include not allowing political speech that incites violence to go unchecked, with quicker escalation of content to human review and tough sanctions on repeat abusers being prioritised.
"More comprehensive labelling of AI-generated content is an important first step in this global year of elections, but too much is at stake for companies to drag their feet when it comes to adopting greater transparency across the board," McConnell added.