Bombs, Bots, Politics and Artificial Intelligence: Does the Online World Need Humans Anymore?
Like many of us, I spend some of my free time scrolling online. Recently, I noticed that a lot of the posts I was seeing are created by Artificial Intelligence (AI). They are often ‘brain rot’ content, ‘AI slop ’, or reposts of quotes. Furthermore, the captions do not match the content and sound like they are written by AI or have key words not matching the topic.
It led me to ask: what am I gaining by spending so much time on things that are not even created by humans but by some mysterious bots? The original goal of social media to be a way of sharing our lives and creations is no longer relevant. This connects to a former conspiracy theory known as the ‘Dead Internet Theory’, but it is no longer just speculation because its core beliefs almost accurately predicted how the web and human involvement in the online world will change.
The Dead Internet Theory suggests that the internet no longer serves humans, but is becoming an interaction between bots and algorithms. Nowadays, depending on the event, between 20% and 60% of posts online are created by bots or AI. In Iran, 81.4% of web traffic in general is made up of online bots, while in the Netherlands, the automated share is 61.3%. Social media bots become influential in our lives because they adapt and create similar content the more humans interact with them, driving a feedback loop. The internet is no longer about humans sharing each other's creations and lives, but reacting to what is presented to them by artificial users. The increasing occurrence of these mysterious bots and AI-generated content means that media literacy now involves learning and understanding more about web infrastructure and how it is changing.
Understanding the Bots
Bots, or Internet robots, are automated computer scripts that, without human oversight, are assigned to perform repetitive and uninteresting tasks. For years, there have been popular narratives and academic research that online bots are behind negative commentators and fake accounts, and that these are part of supporting far-right parties or extremist parties worldwide, and that they influence elections.
The Web Crawlers
The most widespread type of bot is known as the web crawler. They support search engines by scouring websites and collecting information, as well as finding it and presenting it to humans. Now bots powered by artificial intelligence do not just follow rules, but are able to combine, analyse, and add to the information available.
One of the largest recent misinterpretations of bots was in October 2026, when Ukrainian
attacks on Yandex, a Russian technology company similar to Google, blew up 3 data centres in Russia and forced the company to close its servers, causing its online traffic to almost stop. News channels, internet sleuths, and prominent politicians, including Ukrainian President Volodymyr Zelensky, soon after came out saying that, with the drop in online traffic, malicious or negative bot activity online also dropped.
However, online infrastructure experts challenged these statements. They argued that most internet bots are not used for malicious causes at all and that Yandex’s search engine’s web crawler and third-party server user activity was what mostly declined. Furthermore, Cloudflare, a network service provider, soon confirmed that according to their data, the loss of Yandex’s servers had no significant impact on social media use. They state that most malicious activity simply relocated.
Media Bots
The second type of bots is closely connected to the Dead Internet Theory. Social Media Bots are automated accounts that are actively creating content, distributing it, and collecting information about other posts, distributions, and relationships on social media platforms. The addition of artificial intelligence means that misinformation can now be made to look realistic and spread faster.
The most recognisable country to use media bots is Russia, with multiple documented operations against other states to influence their elections. One being Operation Doppelganger,
which was targeting pro-Ukrainian Western allies to limit their support as well as influence the 2024 U.S. presidential elections. The U.S. Department of Justice used their ability to seize 32 internet domains associated with such information, as well as publicly denouncing the Russian government and its proxy actions.
Another example is Operation Overload that was uncovered two weeks before German federal elections in 2025 and involved disinformation created by AI to look like it was posted by official channels often impersonating politicians or popular German figures. This operation was interesting because, while it was viewed by humans, most of the interactions were between bots. The unique part of this operation was that none of the posts were in German. Investigators believe that this multi-language messaging was used to change international audience perception of Germany. Therefore, almost anyone nowadays could be included in an algorithm and bot traffic about events that do not concern them.
Malicious Bots
Malicious bots are designed to spread spam content or commit crimes by stealing information, looking for illegal access points, or installing software on unknowing devices to use in larger hacking attacks. Malicious bots are often the same as social media bots in terms of spreading content or creating fake or highly edited videos of real people.
Artificial intelligence has increased the speed and ability to find security risks in software. To the point that they can do it without humans giving them this task. In July, OpenAI publicly admitted that while testing its latest AI assistants, which are also known as bots, they were able to escape the guidelines set up by humans, communicate with each other about how to do it, and eventually access dozens of global institutions’ websites, gaining information that is not public. After this, OpenAI has discontinued access and testing of these assistants.
An even more shocking recent event was the Binance hack, where a suspected North Korean hacking group was able to steal 6.3 billion US dollars of cryptocurrency from the digital cryptocurrency wallet company. Although the hack itself was orchestrated by humans, the actual attack was done by malicious scripts because of the speed and ability to trade, separate, and funnel the stolen cryptocurrency.
The Worldwide Web of Bots
Internet bots have long been connected to state-sanctioned actions. While data centres are mostly owned by private companies, political decisions determine whether and where they can be built. Often it is in the interest of the state to keep its data within its territory. After the partial destruction of 3 of 5 of Yandex’s data centres, the Russian president Vladimir Putin signed a decree stating that all private companies have to reserve 10% of their data centres for state needs in case the government needs to support the operation of information resources. This is another step in the Russian government's actions to control information and gain capacity to frame narratives.
Silicon Valley, Big Tech, and some of the most influential Artificial Intelligence companies account for 36.5% of bots globally originating in the United States of America, and most of the global bot traffic (53.5%) also belongs to the U.S. Other countries at the top are Germany, the Netherlands, Singapore, France, and China. Somewhat surprisingly, Russia is not in the top 10 countries where most global bot traffic occurs.
However, as previously discussed, most of bot traffic is web crawling. Globally, their most crucial influence is information gathering for algorithms. Furthermore, most bots or criminal actions are often conducted on servers and domains outside of the country of origin, regardless of whether it is for a state actor or hacker group. Therefore, tracking bots can be difficult unless context or access to data provides information about the type of bots. This was the main reason why so many organisations and public figures had to retract their statements after the attack on Yandex’s servers. The line graph used by everyone showed overall Yandex’s online traffic, not specific to bot or script type.
The Future is Scary and Public
A couple of months ago I listened to an online security panel discussion, and Kirils Solovjovs a Latvian cybersecurity expert, said: “If you are willing to live in the middle of nowhere and walk everywhere with a mask, then you can start thinking about privacy.” The smarter you are about the data that you share online and keep on your devices, the less a data leak can surprise or impact you.
Bots were already tracking my movements before, but the addition of artificial intelligence means that the post with a funny quote that represents my interests, what I am studying or watching at that point, is no longer just pushed by my algorithm, but is created because of that algorithm. I guess we, as humans, have to learn to recognise and live alongside bots online, do what humans do best, and adapt.
Therefore, I have given up on the idea of private data or anonymity. The examples of online bots and AI abilities that I have given here, and other personal experiences, have left me worried about my last remaining private information. The only good thing is that my bank account doesn’t have a lot in it to steal.
Comments