AI and visual propaganda are being used to stoke tensions as Ethiopia eyes Eritrea's Red Sea Port

Aerial view of the Grand Ethiopian Renaissance Dam (GERD).

Aerial view of the Grand Ethiopian Renaissance Dam (GERD). Image by the Prime Minister's Office, Ethiopia. Public domain via Wikimedia Commons.

By Amanuel Tesfaye and Matti Pohjonen

On September 9, 2025, Ethiopia inaugurated the Grand Ethiopian Renaissance Dam (GERD), a mega-dam with the capacity of generating over 5,000 megawatts of energy. The ceremony was highly publicized, with both state outlets and social media ablaze as Ethiopians celebrated a dam that took 14 years and USD 5 billion to build. On the digital sphere, however, another peculiar phenomenon was happening alongside the celebrations: the proliferation of videos and images generated with artificial intelligence (AI). Some of these videos and images contained a not-so-subtle message about where Ethiopia should focus its efforts next: acquiring the port of Assab, which is held by its neighboring country, Eritrea

Viral AI-generated content

One AI-generated video, widely circulating on TikTok and X, and shared by former State Minister Birhanu M Lenjiso, celebrates the completion of the dam while calling for taking over the Assab port as the next goal. The video starts with a fictional mega-stage simulating a concert at the site of the GERD, a stage adorned by Ethiopian flags, surrounded by a loud synthetic audience. An AI-generated voiceover says:

How joyful it is to be Ethiopian! 

How joyful it is to live in the era of the Renaissance!

How honorable it is to be a maker of Red Sea history! 

May Ethiopia live forever! 

From the Nile to Assab, one people! one heart!

The announcement is followed by AI-generated patriotic music and a video. In the second half of the video, an unmistakable message appears: “GERD 2.0 Coming Soon!” followed by “See you all in Assab!” Find the full AI-generated video here:

AI for political propaganda

A map featuring the horn of Africa and Port of Assab.

A map showcasing the Horn of Africa and the Port of Assab. Copyright Z FreCC BY-SA 4.0, via Wikimedia Commons.

The proliferation of such synthetic propaganda is unfolding amid tensions in the Horn of Africa over Ethiopia’s desire to secure unfettered access to the sea. In 2023, Prime Minister Abiy Ahmed announced that access to the sea is a matter of survival, lamenting that 150 million people cannot live in a “geographic prison.” This statement alarmed neighboring countries which saw the announcement as a threat to their sovereignty and territorial integrity. This has created fears that an inter-state war might be on the horizon, particularly between Ethiopia and Eritrea.

While Ethiopia has repeatedly stated it will pursue its ambition peacefully and diplomatically, its posture has grown more belligerent, with military leaders repeatedly declaring  that the army is ready to answer the call if the government gives it a “duty or responsibility.” Prime Minister Ahmed also reportedly told local business leaders, “We want to get a port by peaceful means. But if that fails we will use force.”

The use of AI for political propaganda is emerging as a major concern in contemporary democratic societies — particularly those in sensitive geopolitical contexts. AI-generated propaganda, sometimes referred to as “synthetic propaganda” or “slopaganda,” involves the use of generative AI tools to create synthetic images, audio, and video for circulation on social media platforms such as TikTok, Instagram, YouTube, X, and Facebook. 

This new genre has become especially popular in the US and Europe in support of populist parties such as Alternative for Germany (AfD) and leaders such as Donald Trump to mobilize their support base and delegitimize adversaries, often using hyperbolic and whimsical images and videos. In much of the world, this is not only taking place under right-wing populist contexts but also influencing geopolitical conflicts. 

Similarly, in Ethiopia, AI is emerging as a key tool as the government and its supporters increasingly turn to visual propaganda to sway the public. The synthetic videos and images regarding Assab illustrate an evolving trend in how the Ethiopian government is amplifying its official discourse. A close analysis of the pictures and videos reveals the visual tropes and audio-strategies that characterize these messages, as well as the underlying narratives behind them. 

Narratives behind the AI-generated content

Visually, the images and videos are crisp and colorful, almost always adorned with the nationalist imagery of Ethiopia's green, yellow, and red flag. They depict ultra-modern ports buttressing an expansive blue sea, ostensibly representing Assab and the Red Sea. Large arch-like signs that depict the location as “ASSAB” feature frequently. Military naval ships, sometimes in exaggerated numbers, are present, marked unmistakably as Ethiopian by the flags they carry. Most of the vessels carry Navy sailors in crisp white uniforms. Some explicitly contain the phrase “no more landlocked” plastered on the sides of large ships. 

Another popular text is “from GERD to Assab,” signaling a desire to ride a wave of development and support for the GERD to rally society to pursue the port agenda next. The Ethiopian army and its mechanized divisions taking over Assab is also a popular imagery. In other instances, the Prime Minister is featured alongside the port, illustrating the importance of the leader in articulating this vision of reclaiming the country’s lost access to the sea. 

The proliferation of short-form videos means the sound accompanying the videos has also emerged as an important instrument of political messaging. AI-generated patriotic songs accompany the visuals. One AI-generated song, for instance, croons: “Assab, my bride, the gateway and passage of my being. Do not despair, my country — the hour has come; your landlockedness will soon be over!” Find the full AI-generated video below:

These videos and images are also accompanied by text amplifying the message either in the form of a caption or embedded within the video/image itself. One caption from a popular pro-government Twitter account reads:

“The Red Sea is in our past, and it must be in our future. For dignity, for development, and for generations to come, Ethiopia’s rightful access to the sea must be restored.

While another, accompanying an AI-generated image of a port with Ethiopian naval ships, declares: “My next summer vacation destination! #Assab #Ethiopia.”

Behind these visual tropes and audio files are deeply ingrained narratives that seek to justify and legitimize Ethiopia’s aspirations to reclaim its lost sea access. AI-generated content functions by appealing to patriotism, aiming to mobilize the nationalist sentiment of Ethiopians. The country’s landlockedness and the manner in which Ethiopia lost its seacoast are perhaps among the few political issues that unite an otherwise fragmented polity. There is a sense of loss and regret, an emotion these videos aim to tap into in an attempt to generate public support for the government’s ambitions.

With the help of AI, these images and videos paint an alternative vision for the future — a future in which Ethiopia has escaped the “geographic prison” of landlockedness and rides on a path to “prosperity.” 

This phenomenon illustrates how AI is being used to heighten and accelerate visual propaganda, as it has become significantly cheaper to produce and disseminate content as a result of the popularisation of AI-driven image, video, and audio generation tools. Crucially, it highlights state actors’ ability to quickly recognize and tap into the changing nature of platform cultures, where short-form video has increasingly supplanted text as the preferred form of content among audiences. Just as importantly, it signals that the classic division between truth and falsity no longer matters, as we enter what some critics have termed the “Fuck It” Era of AI-Generated Slop.

Start the conversation

Authors, please log in »

Guidelines

  • All comments are reviewed by a moderator. Do not submit your comment more than once or it may be identified as spam.
  • Please treat others with respect. Comments containing hate speech, obscenity, and personal attacks will not be approved.