Journalists defend information integrity in the AI age

A guest piece by Linea Dishena and Eyob Abraha

After the formal end of the Tigray war, Ethiopians prepared themselves for the general elections on  June 1, 2026. One week before, an AI video was circulating on Facebook, showing disgruntled queuers and a uniformed, armed man aggressively pushing a frightened woman voter towards the ballot box. The caption assertion was, "Ethiopia #Vote 2026: Abiy Ahmed Ali is threatening to use weapons to force people to vote for him in the 2026 elections and has arrested many people; he is a war criminal." Within minutes, views and reactions had already gathered, sparking speculation and fear before fact-checkers and journalists could determine the video's inauthenticity.

On 08 June 2026, Ethiopia's Menahria Radio 99.1 debunked it as entirely fabricated after it had been circulating for 16 days. Their fact-check states: "Our investigation confirms that the viral video is entirely fabricated. While it looks convincing at first glance, a closer analysis reveals it is an AI-generated deepfake filled with clear digital anomalies."

While AI offers new efficiencies, it has also unleashed a wave of fake content that threatens to overwhelm traditional verification systems, such as official statements, cross-checking, real-time reporting, news sources, and interviews. In Ethiopia, journalists are confronting a wave of polarized AI content on social media, created by political actors. At the same time, there are only limited tools available for verification.

Disinformation spreads rapidly on social media

Long delays in fact-checking are especially problematic in Ethiopia, where conflicts in the Ethiopian Tigray, Oromia, and Amhara regions have been ongoing since November 2020. A humanitarian crisis is ongoing, despite a peace agreement signed in November 2022 by one of the fronts, the Tigray People's Liberation Front (TPLF).

"Most breaking news in Ethiopia, especially on security-related information, first appears on social media platforms such as Facebook, Telegram, TikTok, and YouTube," says Pawlos*, a political and current affairs reporter. "But official confirmation from government offices is often very slow. Sometimes it takes hours or even a whole day to get a formal response."

The time required for fact-checks stands in stark contrast to the speed and frequency with which disinformation spreads. Independent fact-checking organization MultiFact Check's monthly report on Ethiopian disinformation trends and social media monitoring recorded 836 cases of disinformation in Ethiopia in May 2026, up from 709 in April. 

According to the report, TikTok and Facebook remained the primary distribution channels of disinformation. At the same time, Telegram continued to serve as a key space for the sharing of organized political content. The messenger app is considered the backstage, where many coordinated campaigns take shape before going public. "As voting day approached, content actors across the political spectrum accelerated production of misleading and outright false material," the report states.

Challenges abound in Ethiopia's information ecosystem

In a media landscape like that of Ethiopia, shaped by limited press freedom, political tensions and slow official communication, the challenge of AI-driven disinformation becomes even more acute.

Ethiopia ranks low in global press freedom assessment, at 148 out of 180 countries, according to Reporters Without Borders' (RSF) 2026 World Press Freedom Index. "The media landscape under Abiy Ahmed, who came to power in 2018, remains very polarized and is characterized by opinion journalism to the detriment of fact-checking," RSF states.

Digital access remains uneven. A 2025 UNESCO report on internet development in Ethiopia indicates that 25% of the population are regular internet users.

According to the DataReportal's 2026 report, there were 9.8 million social media users in Ethiopia in October 2025, or 7.2% of the population. Digital media such as Facebook, TikTok, and Telegram are playing an increasingly central role in how news is shared.

The burden of verification

In the race to spread and verify misinformation, the fundamental work of journalists often falls by the wayside. Journalists today are no longer simply reporting events; they are first trying to determine whether those events are real. This growing "verification burden" is reshaping newsroom practices and priorities.

 "As I am living and working in a country undergoing state-building, AI-generated content is high, and some political miscalculations have made my people suffer a lot," said Girma*, an assistant editor for news and current affairs.

Amid ongoing political transformation and conflict, social media has evolved into what researchers at the Africa Center for Strategic Studies describe as an "information war," in which competing narratives are rapidly produced and amplified. In such a context, verifying information is not only a technical task. It happens in a heated political climate.

For fact-checker Tesfaye*, the additional work to verify information is profound. "Journalists' work is made harder due to the spread of false information," he explains. "More time is spent on verification rather than producing original content."

A challenge and a much-needed tool

There is another perspective on AI in journalism. The same technology that accelerates disinformation is also expected to help counter it.

In Ethiopian newsrooms, journalists are experimenting with AI tools for transcription, data analysis, and elements of fact-checking, according to the Reuters Institute for the Study of Journalism, while verification relies to a large extent on metadata analysis and reverse image searches.

"We are constantly adapting, and AI is a huge help for fact-checking," says Abera*, a deputy editor-in-chief. However, getting hold of the appropriate tools and managing the volume of fake news at the same time is difficult: "The tech is just moving so fast. It's getting so sophisticated that it's hard to even track where the content came from," continues Abera. "Then there's the sheer volume of it during a political crisis; there is so much content flying around that it's overwhelming for journalists to check it all. A lot of traditional media houses here also just don't have the budget or the tools to do this kind of analysis."

In practice, this leaves journalists navigating a constant trade-off: Publishing quickly and risking the spreading of falsehoods, or verifying thoroughly and falling behind on the news cycle.

Journalists taking their own initiative to build AI skills

Increasingly, they are left to make these calls on their own. Here, in the same breath, deputy editor-in-chief Abera acknowledges that his team lacks the skills and access needed to use the tools fully, noting that adaptation often happens informally and is unevenly driven more by individual initiative than by institutional strategy.

"We have limited access to the technology tools themselves, and our team needs more training," Abera added.

Unlike, for example, the BBC, Ethiopian media organizations have not yet developed editorial guidelines for the use of AI in content production. "There are definitely gaps," Abera said.

In the meantime, journalists are not waiting for systemic solutions; they are teaching themselves, experimenting with tools, and trying to keep up with rapidly evolving technologies. Many rely on online tutorials and self-study to build AI skills, highlighting both initiative and a lack of support.

"I did not get any training or attend any training, but personally, I read a lot of things related to AI tools to further my knowledge," said Nahom*, a former journalist at a state-owned media outlet.

Many of the current AI tools have been developed outside Ethiopia and do not function effectively in many of the languages used in Ethiopia. As a result, detecting AI-generated disinformation remains difficult, says Pawlos, the political and current affairs reporter at a broadcasting corporation. "We need AI systems that are trained properly in Amharic, Afaan Oromo, Tigrinya, Somali, and other Ethiopian languages," he adds.

When responsibility comes with risk

When false information spreads unchecked, the consequences are not only professional but also personal.

 "When misinformation circulates online, and we cannot respond immediately, some people blame government media for not correcting it fast enough. I received threatening messages from armed groups and individuals who believe the false information," Pawlos says. Threats and other pressures can restrict journalists' ability to move freely and report safely.

According to fact-checker Tesfaye, who mainly focuses on misinformation connected to the current conflict in Ethiopia, journalists may hesitate to challenge information promoted by powerful actors, fearing retaliation: "There is a threat to journalists who try to publish fact-checked information that goes against the political agenda set by the government or the opposition," he says. In this environment, verification is no longer just about accuracy; it is about safety and navigating power structures.

"It is up to us."

Faced with these pressures, Ethiopian journalists are increasingly defining their role not just as reporters but as active defenders of information integrity. "A journalist is as equal as a soldier. You can die anytime; you can be arrested anytime," said Meriam*, who has worked as a content producer for more than eight years.

In the race against AI-driven disinformation, Ethiopian journalists are not only catching up with technology; they are redefining what it means to do journalism when verification, safety, and trust are constantly under threat. Thus, alongside the pressure for speed — and in times of conflict and political actors spreading misinformation — they are often left to fend for themselves. "I work under an anonymous name because I don't want to be arrested," said Yonas, who has worked as journalist and fact-checker for six years.

In this environment, journalists are adapting under pressure, learning new technologies without formal support, and navigating risks that extend beyond the newsroom. To live up to their responsibilities, they are calling for something more: access to tools, training, and systems that match the scale of the challenge they face.

But for now, however, much of the responsibility remains with them.

*We have used pseudonyms for all interviewees in this article to protect their identities.

Lead author: Linea Dishena is a multidisciplinary journalist from Namibia with eight years of experience in print, broadcast and digital media. She holds a Master's degree in International Relations and Diplomacy and is currently pursuing a Master's in International Media Studies at DW Akademie.  

Co-author: Eyob Abraha is a media professional from Ethiopia with a background in media and technology and experience in media production, TV and strategic communications. He is currently pursuing a Master’s in International Media Studies at DW Akademie. His interests include how emerging technologies are reshaping journalism and public discourse.

Editorial support: Mira Keßler

Research support: 
Tracy Naeku Memusi
Francesca Amofah
Janice Bockstiegel
Aditi Shrestha
Sophie Steiner