AI Promises in Ethiopia Shadowed by Deepfakes and Restricted Internet Freedoms, says New Report

A new report by the Collaboration on International ICT Policy for East and Southern Africa (CIPESA),a think tank, warns that while AI could streamline governance and service delivery for millions, it is also amplifying misinformation, deepfakes, and state surveillance in one of Africa’s least connected nations.

With internet access reaching only 21% of Ethiopia’s 126 million people, the lowest among 14 African countries surveyed, CIPES, the research, advocacy and capacity-building organization working at the intersection of technology, policy, rights, and governance in Africa, describes a “digital future” for Ethiopia unfolding amid restricted freedoms, limited connectivity, and a growing flood of manipulated media.

The promise of AI is evident across sectors. Yet the dangers are mounting just as quickly. Deepfake videos of athletes Haile Gebrselassie and Kenenisa Bekele claiming they had launched a mobile app were circulated on TikTok and Facebook in January 2025. Also, cloned voices attributed to Prime Minister Abiy Ahmed (PhD) were circulated on social media.

Notably, the State of Internet Freedom in Africa 2025 report states synthetic news articles and algorithmic amplification on Facebook contributed to hate speech and incitement to violence during the Tigray conflict in 2020-2022. “The content was used to intimidate journalists and activists, particularly in conflict-prone regions such as Amhara and Oromia. Experts have warned that emerging tools could fan ethnic divisions and erode trust in democratic processes,” reads the report. 

According to the 76-page report, these incidents highlight the risks of algorithmic bias in AI models trained on foreign data, systems that misinterpret Ethiopian languages and contexts, reinforcing exclusion for women, rural citizens, and ethnic minorities. At the same time, an ongoing “AI brain drain” is depleting the local talent pool, leaving institutions dependent on imported technologies.

Meanwhile, Ethiopia’s internet environment remains tightly controlled. The country scored 27 out of 100 on Freedom House’s Internet Freedom Index, classified as “Not Free.” 

The government has taken notable steps toward regulation. In July, it ratified a National AI Policy and endorsed UNESCO’s ethical framework, followed by the release of a second national AI strategy extending to 2030. The framework aligns with the African Union’s broader digital transformation vision and emphasizes responsible use of emerging technologies.

Ethiopia’s latest federal budget, ratified in July, allocates 1.13 billion Birr to the Ethiopian Artificial Intelligence Institute (EAII), signaling the government’s growing interest in a technology reshaping the global economy. The figure represents a 42% increase from the last fiscal year’s budget of 795 million Birr to the three-year-old institute

Universities and research institutes have also begun holding workshops and multi-stakeholder forums to discuss ethical AI deployment. The Digital Ethiopia 2030 blueprint envisions technology as the backbone of a growing economy, which expanded 6.6% last year despite inflation and political strains. But as the CIPESA report makes clear, realizing that vision will require more than infrastructure and innovation. It calls for stronger regulation, greater transparency, locally built datasets to counter bias, and literacy campaigns to equip citizens for an AI-driven era.