Do not have AI verify the six-fingered Netanyahu

There are periods when truth disappears. But even more dangerous are the periods when truth is drowned out by fake images, videos, and speculation. This is exactly what is happening on social media today during a time of war, bombing, missile attacks, assassinations, blockades, and shutdowns. On one side, there are fake images produced by AI; on the other, algorithmic authority, oversight, censorship, and marginalization, where certain accounts are shut down or rendered invisible. Much like in the film The Trial (1962), the rule of invisible algorithms on social media reminds one of characters in a Kafkaesque order who are punished without knowing what their crime is. Social media platforms' “sensitive content” policies are often defended on the grounds of security and ethics. However, in practice, these filters are also an algorithmic regulatory tool that limits the circulation of certain political narratives. Accounts and posts that do not align with or oppose the political economy of technology companies are made invisible. Under the label of sensitive content, what is legal and legitimate is marginalized. Especially on X, many anti-imperialist, anti-capitalist, and anti-Zionist individuals, groups, and party accounts are trying to exist on the platform under an algorithmic fog.

On the other hand, some videos circulating on social media in recent days have triggered a new debate: manipulations and claims regarding the death of Israeli Prime Minister Benjamin Netanyahu, and whether the images shared to refute this were produced by AI. Claims of AI manipulation were raised based on details in the images such as six fingers, facial features, and ear expressions. These videos may or may not have been produced by AI. But the debate itself reveals another, more fundamental problem: in a digital media landscape where we can no longer easily trust any image, what is the truth, and how is it to be found and shared? The issue is actually more dire in this respect: AI is not only producing fake images; it is also increasing the suspicion that real images might be fake. Thus, even images of war turn into a pixel-examination debate within minutes, and the importance of truth is lost, becoming tabloidized within technical details and memes.

The issue of the fragmentation of reality is also a very old debate in cinema. In the 1952 masterpiece Rashomon, Akira Kurosawa shows how the same event is told in completely different ways by different people. Each witness produces a different truth according to their own interests and perspective. There is a similar Rashomon effect in today's social media. The same image can turn into dozens of different narratives within a few hours. In the digital world, AI fact-checkers can select only the one “most suitable for the ownership structure” as the absolute truth and throw the rest into the digital media trash heap by calling it “sensitive content” or “disinformation.” We do not watch reality, but the image presented by the algorithm with class-based selectivity.

The main character in the film The Truman Show (1998) only realized he was living in a fake world when he hit the boundaries of that world. Our verification and truth-checking mechanisms are like the set walls at the end of the sea that Truman tried to escape. The “cannot be verified” or “taken out of context” warnings we encounter when we question Israel's aggression or the hypocrisy of Western media may actually be the painted sky that Truman's boat hit. Tools like Groke or similar ones seem to want to tie us to a port owned by digital technology companies, telling our boat that is searching for the truth, “stay in this simulation of reality that we have approved.”

In the film Wag the Dog (1997), a war was staged in a television studio to distract the public, while the question of what the difference is between a human and their copy, once debated in science fiction cinema, was explored through androids in the film Blade Runner (1982). Today, we ask AI to distinguish between real/fake/artificial humans. In his novel “Blindness,” Jose Saramago describes the conscientious blindness and moral collapse of a society that does not want to see the truth even though its eyes can see. While we argue over a video about whether Netanyahu is dead or not, like the blind people in Saramago's book, we cannot see the destruction in the war and the deaths of hundreds of people. Our need to have things verified by AI turns us into voluntary blind people who do not trust our own eyes. While the algorithm drowns us in a noise where technical details are debated, the real tragedy that our conscience should see—the goals and objectives behind the wars—escapes our notice.

In the film The Matrix, Morpheus tells Neo, "What is real? If you're talking about what you can feel, what you can smell, what you can taste and see, then 'real' is simply electrical signals interpreted by your brain." Today, reality is turning into algorithm signals coming from the owners of the platforms. On the other hand, Étienne de La Boétie argued in the 16th century that the power of tyrants feeds on people's habits and comfort. With the concept of “Voluntary Servitude,” he points out that people become open to manipulation because they avoid the political responsibility that comes with facing the truth. In this trajectory, the user, in their cognitive comfort and verification laziness, first settles for content that does not disturb their own views. Later, instead of researching or waiting to see if a video is real, they leave the verification work to another AI within the logic of social media, which is built on interaction and speed. AI and verification algorithms, in turn, present what passes through the “truth” filter permitted by the ownership structure as reality. Thus, the responsibility of questioning the truth is once again delegated to technology. In this context, the way out of Anadolu Agency's glorification of F-35s, X's ownership-focused censorship, and the visual chaos created by AI requires a critical political-economic perspective that looks beyond the sky drawn by the algorithm.

A simple but important question needs to be asked here: If the system that produces and circulates an image and the system that verifies it are part of the same technology ecosystem, how independent can this verification be? In short, can the technology where manipulation is produced or can be produced also be the tool where the truth is verified? Because the de facto owners of the black-box algorithms that create the filters for verification processes are also the global technology companies, mostly based in the US, that control the economic and technical infrastructure of digital platforms.

For this reason, do not have AI verify the “six-fingered Netanyahu” that is claimed to have been created with AI. Due to the deep learning technique based on competing artificial neural networks, the closeness of images created by AI to reality will increase day by day. New manipulations will be produced. New verification tools will be released to the market. But the question will remain the same:

Are we really looking for the truth, or do we prefer to live with and view life through the version presented to us by the algorithm?