Addressing misinformation and deepfakes

Sir Arthur Conan Doyle believed two schoolgirls had captured photographic evidence of fairies playing in a garden in Yorkshire. (Photograph taken by Walter Bennington in 1914, via Wikimedia Commons)

When I was a boy, our schoolteacher showed us some photos of fairies purportedly playing with two young girls in a back garden in Yorkshire. The pictures, taken between 1917 and 1920, certainly had me fooled and I suspect many of my classmates too. <_o3a_p>

You are probably thinking that five-year-olds are not the most discerning audience but, be that as it may, we were in good company. Among those who had been taken in by the so-called Cottingley fairies was none other than the creator of Sherlock Holmes, Sir Arthur Conan Doyle. <_o3a_p>

Indeed, by all accounts there was a lively public debate in 1917 and to settle the matter somebody sent the photographs to the Kodak company for a professional opinion. Unfortunately, Kodak's forensic experts could find no evidence that they were fakes and many people - including grown men and women - continued to believe in what they saw as compelling evidence of supernatural phenomena.

<_o3a_p>AI and Multimedia standards collaboration

Fast forward to 2025 and we find ourselves in a similar situation. Nowadays, though, it is not pictures of fairies created by bored schoolgirls but AI-generated deepfakes that are confusing and dividing public opinion.

Today's fake media images are far more dangerous and just as difficult to debunk as the pictures from Cottingley. International standards offer hope and one in particular, known as JPEG Trust, is already a powerful tool for authenticating certain kinds of content, but more are needed.<_o3a_p>

For this reason, the World Standards Cooperation (IEC, ISO, ITU) set up a global, multistakeholder initiative last year to address the challenge. The AI and Multimedia standards collaboration brings together standards developers, technology leaders, policymakers and civil society to combat deepfakes, misinformation and the misuse of AI-generated media. <_o3a_p>

In the framework of the collaboration, technical experts are currently mapping the standards landscape. Their task is to identify gaps where standards are needed but not currently available to support the policies that governments, industries and other stakeholders want to implement. <_o3a_p>

At the same time, policy experts are focusing on the interplay between government policies and standards, as well as identifying best practices to inform policy development.<_o3a_p>

A third goal of the collaboration is to educate stakeholders about the scale of the challenge and to shine a light on how different sectors are fighting back and the role of standards. <_o3a_p>

In the upcoming webinar Addressing misinformation and deepfakes in the media, distinguished experts will examine how newsrooms, social platforms and civil society organizations combat fake media. They will look at real-world consequences and the tools, policies and standards needed to restore trust in online content. <_o3a_p>

Why this matters now more than ever <_o3a_p>

In 1917, the Cottingley Fairies were a curiosity. Today, deepfakes can swing elections, incite violence and destroy reputations in seconds. <_o3a_p>

The stakes have never been higher. <_o3a_p>

Join the interactive webinar on 30 April 2025 at 14:00 - 15:00 CEST Geneva: register to receive a reminder or watch live.