IN S FA ES TE N T HA RM Disinformation is content that is intentionally false and designed to cause harm. It is motivated by three distinct factors: to make money; to have political influence, either foreign or domestic; or to cause trouble for the sake of it. EN TO LS D I S I N FO R M AT I O N , M I S I N FO R M AT I O N A N D M A L I N FO R M AT I O N Misinformation Disinformation Malinformation When disinformation is shared it often turns into misinformation. Misinformation also describes false content but the person sharing doesn’t realise that it is false or misleading. Often a piece of disinformation is picked up by someone who doesn’t realise it’s false, and shares it with their networks, believing that they are helping. The sharing of misinformation is driven by sociopsychological factors. Online, people perform their identities. They want to feel connected to their “tribe”, whether that means members of the same political party, parents that don’t vaccinate their children, activists who are concerned about climate change, or those who belong to a certain religion, race or ethnic group. The third category we use is malinformation. The term describes genuine information that is shared with an intent to cause harm. An example of this is when Russian agents hacked into emails from the Democratic National Committee and the Hillary Clinton campaign and leaked certain details to the public to damage reputations. We need to recognise that the techniques we saw in 2016 have evolved. We are increasingly seeing the weaponization of context, the use of genuine content, but content that is warped and reframed. As mentioned, 8 I N F O R M AT I O N D I S O R D E R anything with a kernel of truth is far more successful in terms of persuading and engaging people. This evolution is also partly a response to the search and social companies becoming far tougher on attempts to manipulate their mass audiences. As they have tightened up their ability to shut down fake accounts and changed their policies to be far more aggressive against fake content (for example Facebook via its Third Party Fact-Checking Project2), agents of disinformation have learned that using genuine content — reframed in new and misleading ways — is less likely to get picked up by AI systems. In some cases such material is deemed ineligible for fact-checking. Therefore, much of the content we’re now seeing would fall into this malinformation category — genuine information used to cause harm. INTRODUCTION 9