Twitter's role in modern warfare
Virtual warfare is being waged on social networks with the majority of users unaware that nations and militant groups alike are targeting their hearts and minds.
Modern conflicts are no longer about warring states and control over territory, but more about identity, control of the population and the political decision-making process, argues military researcher Thomas Elkjer Nissen.
This makes Twitter - which is marking its 10th anniversary - ripe for exploitation in any conflict. It is the network users turn to for news, and where news organisations break stories, so interested parties need their version of events to appear on timelines
It has been used most notably by jihadist groups but, increasingly, there are also worries that states such as Russia may be using the network to influence populations without anyone noticing.
The number of supporters of the so-called Islamic State on Twitter is tiny - about 46,000 in total, according to a Brookings Institute survey in 2014.
Despite this, the group has used Twitter to gain and hold the attention of a mass audience, and to attract supporters.
Researcher JM Berger of George Washington University has laid out how the group "grooms" new members: from looking for potential supporters in Muslim-oriented networks, to surrounding targets with a small community that interacts with them and then encouraging a recruit to take action.
"The user often starts with a link in a tweet and is then led further and further into the narrative on blogs, videos and other social media - possibly ending up in a conversation with a recruiter," says Mr Nissen.
"As one of the big three open social networks used by terrorists - Facebook and YouTube being the other two - Twitter is a quick way to send messages, which are easily redistributed by supporters, linking across different media."
Keeping it short
Just ahead of its 10th birthday, Twitter's chief executive Jack Dorsey confirmed it would keep the 140-character limit of its tweets.
"It's a good constraint for us and it allows for of-the-moment brevity," he told NBC's Today show.
The firm had previously indicated it was considering extending the limit to 10,000 characters, or about 2,000 words.
The proposal had met with resistance from its users, with many arguing the restriction was integral to Twitter's identity and that the firm was pandering to its investors, who wanted faster growth.
Amplifying and inflating
IS communicates very effectively on Twitter. It puts out timely information in several languages and uses a variety of multimedia, from cat images to horrifying killings, to engage its audience emotionally.
When IS captured Mosul in Iraq in 2014 the group aggressively used bots and spammed popular hashtags to ensure its propaganda was visible on Twitter, even to people not looking for it.
Berger has estimated that as many as 20% of tweets from IS supporters could have been created by bots or apps.
And IS is not the only group to use Twitter for propaganda. When Somali terrorist group Al-Shabab attacked the Westgate Mall in Nairobi in 2013, the group crowed over the attack and live-tweeted events, easily creating new accounts every time one was shut.
Such events can create the impression the group is stronger and has more support than it actually does. While IS's tweets are easily identifiable, some believe a greater threat may come from the anonymous use of Twitter by states.
Several Russian and Western media organisations have documented the experiences of former Russian trolls and their work and reported observations of suspicious behaviour on social media.
There is still a lack of hard evidence linking such activity directly to the Kremlin. The covert nature of trolling makes it difficult to estimate the extent of influence Russia's trolls may have on Twitter.
Researching this field is social science student Lawrence Alexander, who maps relationships between accounts and has identified what he believes are pro-Kremlin Twitter bots promoting Russian news agencies and pro-Kremlin blogs.
In one instance, he found 2,900 accounts tweeting a specific phrase saying that the opposition leader Boris Nemtsov, who was murdered last February, was killed by Ukrainians because "he was stealing one of their girlfriends".
While not completely certain they were all bots, many of the accounts were following each other and a majority had never favourited a tweet, arguably suspicious behaviour when compared with a random selection of accounts.
Troll influence has extended beyond Russia's borders. In the US, Russian trolls are suspected of tweeting false news of disasters in 2014: an accident at a chemical factory in Louisiana and an outbreak of Ebola in Atlanta.
Eliot Higgins, the founder of Bellingcat, a company that crowdsources information about the Syrian and Ukrainian conflicts, found himself a target, particularly over Bellingcat's investigation identifying the missile launcher said to be responsible for shooting down flight MH17.
"I started off by posting a lot on the Guardian live blog comments before I started my own blogs and some of the people from that followed me on to Twitter and still disagree with me strongly and vocally there up until today, five years later," he says.
"What's been interesting for me is having this Syria community of trolls and the community of pro-Russian trolls that built up around MH17 and my work, now coming together after Russia's involvement in Syria. It's nice to bring people together, even when it's in their mutual and obsessive hatred of one person.
"Recently we've even had the Russian Ministry of Defence and Russian Ministry of Foreign Affairs putting out statements attacking Bellingcat. They seem to be basing it on what the trolls are saying," he adds.
Countries with Russian minorities - Germany and the Baltic states - have been following the issue closely and both the EU and Nato have formed units to counter what they see as a propaganda war online.
As a study at the University of Warwick has found, false rumours on Twitter take much longer to resolve than those eventually proven to be true. Or, as the meme would have it: "The amount of energy necessary to refute bull is an order of magnitude bigger than to produce it."
The lesson is: be wary of what you read on Twitter.