Fake social media accounts are not always automated. In many cases, human operators manually run fake accounts to achieve similar goals, often done by coordinated teams managing a set of accounts. In a few cases, we also found evidence of “cyborg” accounts, whose operators combine automation—to drive volume and speed of activity—with elements of human curation, to make them appear to be legitimate accounts. These hybrid accounts can be the hardest to detect and shut down, since they involve elements of genuine human interaction. While we only found explicit evidence of hybrid accounts being used in 9 of the countries in our sample, we suspect that this activity is much more prevalent in practice, given the still-limited data about the production and use of fake accounts. Of the 46 countries in our sample operating fake accounts, the most common strategy involves automation. Much of this automated fake account activity is restricted to the platforms that make automation easy—usually Twitter—and there was evidence of 38 countries having automated systems for generating content and interacting with human users. Human operators are used to manage the fake accounts deployed by governments and political parties in 33 countries. In countries where fake account activity is high, we can safely say that both automation and human operators drive disinformation. Communication Strategies Cyber troops use a variety of communication strategies to disseminate computational propaganda over social media platforms (Table 3). They create their own content, including fake videos, blogs, memes, pictures, or news websites. These content strategies involve more than simply posting forum comments or replying to genuine user posts, but instead are important sources of junk news, and conspiratorial or polarizing information that can be used to support a broader manipulation campaign. Content strategies also involve the malevolent takedown of legitimate content or accounts. In addition to amplifying certain messages, cyber troop teams use content strategies to suppress voices online. Increasingly, we are identifying human-operated and automated accounts being used to falsely mass-report legitimate content or users so that their accounts and posts are temporarily (and mistakenly) removed by the social media channel. In Armenia, China, Ecuador and Russia we found evidence of this malevolent reporting of content to stifle individual expression and limit the spread of content online (Table 3). Increasingly, government actors are creating their own applications, portals or task forces to combat the threat of fake news and foreign influence operations. In some cases, these task forces focus on fact-checking information that is shared across social media, or they allow 12