3
0
Have you ever ventured into a comment section on a political post? It can be a dark place. Once you manage to find an educational post amongst the mounds of ‘brainrot’, only to be met with a myriad of comments hurling racist, sexist, and xenophobic slogans. It can feel like society is deteriorating. However, emerging studies from several NGOs and journalistic outlets are revealing many of these comments are originating from AI generated bots. So, whilst brat summer may be over, a new technological and perhaps far less friendly trend is taking over.
The existence of bots and their use for malicious content is not necessarily a new phenomenon. The New York Times reported in 2017, that Russia was employing ‘trolling factories’ to interfere with the US democratic system. What is new however, is the bot’s increased intelligence and influence, which is on track to surge with Meta’s deregulation on fact checkers. With little to no constraints, these non-human actors are gaining a greater monopoly in the digital realm, placing the integrity of information at existential risk.
What is the evidence of bots and their influence?
During the UK general election campaign, multiple outlets investigated and discovered bot-like accounts operating on platforms such as X (formerly Twitter). NGO, The Global Witness, reported finding 45 bot-like accounts increasing their activity from May 22nd until July 4th 2024. During this time, they produced around 440,000 posts, which amassed over 3 billion impressions.
What is perhaps more alarming than this fact, is once these accounts had achieved their so-called mission, they transitioned to promote new divisive political content, such as climate denial and ‘Great Replacement Theory’ (a racial conspiracy theory).
These accounts appear to have no loyalty code, as soon as the political trends move on, so do their content.
No evidence was found by any of the outlets that any UK political party were using these bots to amplify their election odds. This begs the question of who is behind these bots and their malicious content.
The growing evidence suggests hostile actors such as Russia are responsible for the growth in AI-generated disinformation online. The US Justice Department, in collaboration with foreign intelligence agencies, have reported this by cracking down on so-called bot farms produced by Russian actors. However, it would be premature to blame all malicious bot activity on Russian hostility. The sheer number of bots suggests multiple culprits, but their evolving nature makes it difficult to discern the perpetrator on the other side of the screen. What is apparent to all, is that these technological foes are part of, and partially in charge of our political discourse.
Are robots taking over the world?
The big question in the AI debate is whether it will exceed human intelligence and take over our roles/jobs. This is often spoken about as if it is a distant, when or a what if scenario. Yet, we display a blatant indifference to bots circulating on our social media platforms. These bots not only have tricked most into believing they are real people, but also manipulate our political rhetoric and beliefs. ‘AI is taking over the world’ is not a futuristic philosophical discussion; it is a reality.
The strategy of these bots is simple, “harass, mislead and polarise” . This article cannot assert that this is the reason for a growing apathy towards politics and a notoriously bad year for incumbent parties, but it certainly does not ameliorate it.
Attempts to tackle this issue will be significantly hindered by Zuckerberg’s paradoxical ‘free speech’ initiative, that removes fact checkers and replaces it with ‘community notes’. Allowing for users to be the ones to add caveats and context, which many, as we’ve established, are malicious bots. A move likely done to impress the incoming Trump administration and maintain edge with his competitor, owner of X, Elon Musk.
What can be done?
For now, we must rely on our critical judgement and tell-tale signs to discern what is a bot comment versus a human comment.
Some general indicators include: no profile picture, no followers, no friends, and the appearance to post repeated political content. The content will be posted at a scale and rate that most users would not be capable to do. They will often vary depending on what is particularly controversial or relevant at the time.
Whilst this article may seem rather pessimistic, one thing it signals is we may not be as divided as it seems online.
Average Rating