Tech News Summary:
- Social media platforms are using artificial intelligence to remove harmful and illegal content, but lack the nuance needed to identify human rights violations.
- Graphic war footage documenting attacks on civilians in Ukraine was quickly removed from platforms like Meta and YouTube, despite their public interest exemptions for such content.
- Human rights groups are calling for a formal system to securely collect and store removed content, in order to prevent evidence of war crimes from disappearing.
In recent news, it has been revealed that artificial intelligence (AI) is being used by social media platforms to delete evidence of war crimes from their databases. This revelation has sparked outrage among human rights organizations and sparked conversations about the ethics of using AI to erase evidence of crimes against humanity.
According to a report published by the International Federation of Journalists (IFJ), many social media platforms are now using AI to automatically delete content that violates their policies. While this may seem like a positive development, the report argues that AI algorithms designed to detect and remove inappropriate content are often programmed to prioritize the removal of “offensive” posts over those that document human rights abuses.
The result, according to the IFJ, is that social media platforms are inadvertently becoming complicit in the erasure of evidence of war crimes. Without proper oversight and intervention by human moderators, the use of AI algorithms to delete content could result in the permanent deletion of crucial evidence that could be used to hold perpetrators accountable for their crimes.
The report highlights several examples of this phenomenon, including the removal of videos documenting war crimes in Syria and the deletion of posts discussing human rights abuses by the Chinese government. In some cases, the content was deleted so quickly that it was impossible to retrieve or document.
The use of AI to moderate content on social media platforms is not a new phenomenon, but this report shines a light on the potential dangers of relying too heavily on automation to police the internet. Human rights organizations are now calling on social media platforms to implement more comprehensive moderation systems that prioritize the identification and preservation of evidence of war crimes and other human rights abuses.
The IFJ report concludes by calling on social media companies to take responsibility for the role they play in documenting human rights abuses and to ensure that their moderation policies and algorithms are designed with this responsibility in mind. Failure to do so could result in the permanent loss of crucial evidence and could leave victims of war crimes without the justice they deserve.