In Senegal, we have a saying, 'Ndank ndank mooy jàpp golo ci ñaay,' which means, 'Slowly, slowly catches the monkey in the bush.' This wisdom, passed down through generations, often speaks to patience and perseverance. But today, as I look at the rapid advance of artificial intelligence across our continent, I see another meaning: how powerful forces can quietly, almost imperceptibly, gather what they need, piece by piece, until the full picture of their intent is revealed. This is a story about people, not algorithms, and the silent scramble for Senegal's digital soul.
For months, whispers have circulated in our tech hubs, from the vibrant incubators of Diamniadio to the university halls of Cheikh Anta Diop. Young developers, their eyes bright with ambition, spoke of lucrative contracts for data labeling, transcription, and image annotation. At first glance, it seemed like opportunity, a chance for our youth to participate in the global AI economy. But as I dug deeper, following the threads of these contracts, a more complex and unsettling pattern emerged. It became clear that these small, seemingly innocuous tasks were not just about creating jobs, but about feeding an insatiable hunger for data, data that is uniquely Senegalese, uniquely African, and increasingly valuable to the global AI arms race.
My investigation began with a tip from a former employee of a local data processing firm, a young man named Omar, who asked to remain anonymous for fear of reprisal. He showed me internal documents, mostly anonymized, detailing contracts with shell companies that, when traced, led back to larger, well-known technology corporations. "We were told it was for 'research and development,'" Omar confided, his voice barely above a whisper. "But the scale, the detail they wanted, it felt like more than just research. It felt like building a mirror of our lives, pixel by pixel, word by word." The documents, which I have reviewed extensively, outlined requirements for collecting dialectal variations of Wolof, Pulaar, and Serer, specific visual cues from Senegalese street scenes, and even behavioral patterns observed in locally generated video content. The sheer volume of data requested was staggering, often involving hundreds of thousands of hours of audio and millions of images.
The evidence points to a sophisticated, multi-layered operation. Large tech companies, often based in the US or China, do not directly contract with individuals here. Instead, they use a network of intermediaries: smaller, often newly formed, local companies or regional data brokers. These intermediaries then subcontract the work to an army of young, eager Senegalese workers, often students or recent graduates, who are paid modest wages for their efforts. The workers, many unaware of the ultimate destination or purpose of the data, are simply grateful for the income in a challenging economic landscape.
One such intermediary, 'Dakar Data Solutions,' a company that appeared seemingly overnight, boasted of its "African expertise in data annotation." Their website, now taken down, once proudly displayed partnerships with unnamed "global leaders in AI." Through corporate registration documents, I connected Dakar Data Solutions to a holding company registered in a tax haven, which in turn had ties to a major Chinese technology conglomerate known for its aggressive AI development. This conglomerate, while not directly named in the local contracts, benefits immensely from the data gathered.
When I sat down with Dr. Aïssatou Sow, a renowned Senegalese computer scientist and advocate for data sovereignty at the Université Cheikh Anta Diop, her eyes lit up when she told me about the implications. "This is not just about collecting data, Fatimà," she explained, her voice firm. "This is about building foundational models that reflect their understanding of the world, but with our faces, our voices, our culture. When these AI systems are deployed globally, they will carry biases, but also capabilities, derived from this data. Who controls the narrative then? Who benefits from the predictive power?" Dr. Sow has been vocal about the need for Africa to develop its own AI infrastructure and ethical frameworks, urging for greater transparency from international players. "We must ensure that our digital future is not merely a reflection of someone else's algorithms," she stated, a sentiment echoed by many local experts.
Indeed, the cover-up, if one can call it that, is less about outright denial and more about deliberate obfuscation. The global tech giants maintain that they adhere to international data privacy standards and that all data collection is consensual and anonymized. However, the labyrinthine contractual agreements and the use of multiple intermediaries make it incredibly difficult to trace the data's journey or to hold anyone directly accountable for its ultimate use. When I approached a representative from one of the implicated global tech firms, they provided a boilerplate response about respecting local laws and empowering local talent, without addressing the specifics of data provenance or end-use. "Our partnerships are designed to foster innovation and economic growth in emerging markets," a spokesperson for a major US-based AI company, who wished not to be named, told me via email. "We prioritize ethical data practices and collaborate with local entities to ensure compliance and mutual benefit." This response, while polished, sidestepped the core questions about control and ownership.
This quiet extraction of data has profound implications for Senegal and for emerging nations across Africa. Firstly, it means that the very fabric of our society, our cultural nuances, our linguistic richness, is being digitized and fed into AI models developed by others. These models will then be used to create products, services, and even policies that will impact our lives, from healthcare diagnostics to educational tools, from financial services to surveillance systems. If these models are built on data that is not fully understood or controlled by us, what does that mean for our autonomy? As the Wolof proverb says, 'Ku nekk ak sa dund ak sa dund ak sa dund,' meaning, 'Everyone has their own life to live,' but in the digital age, our lives are increasingly intertwined with algorithms we don't own.
Secondly, there's a significant economic cost. While some individuals are paid for their labor, the true value of this raw data, once refined and integrated into powerful AI models, is immense. It is the digital oil of the 21st century, and we are, in many ways, selling it at a discount, without fully understanding its worth or its potential. We are becoming data suppliers, rather than data owners or innovators.
Finally, this situation raises critical questions about digital sovereignty and national security. In an era where AI can influence elections, predict social unrest, and even guide military strategy, the control over foundational data becomes a matter of national interest. The US, China, and the EU are openly competing for AI supremacy, and our data is a crucial battleground. As Reuters often reports, the global AI arms race is intensifying, and the scramble for data is at its core. We must ensure that our digital assets are not simply leveraged for the benefit of distant powers. The insights gathered from our daily lives could be used to create highly localized, persuasive AI agents, for example, or to develop advanced recognition systems that could be deployed without our consent.
What can be done? The path forward requires a concerted effort. We need stronger data protection laws, not just copied from Europe, but tailored to our unique cultural and social contexts. We need greater transparency from all companies operating within our borders, demanding clear disclosure of data collection practices and end-uses. We must invest in our own AI research and development, fostering local talent and building our own models that serve our people first. Initiatives like the African Union's AI strategy are a good start, but implementation needs to be robust and swift. As MIT Technology Review highlights, developing local AI capabilities is crucial for equitable global progress.
The story of Senegal's data is not unique; it is playing out across many emerging nations. It is a stark reminder that in the digital age, true independence also means digital independence. We must be vigilant, asking tough questions and demanding answers, so that the future of AI in Senegal is built by us, for us, and not just another chapter in someone else's grand design. The time for ndank ndank in understanding this challenge is over; now is the time for decisive action. We cannot afford to let our digital heritage be quietly siphoned away, leaving us as mere spectators in the AI revolution. We must claim our seat at the table, not just as data providers, but as architects of our own digital destiny. OpenAI and other leading AI labs are constantly pushing the boundaries of what's possible, and we must ensure that our voice is heard in shaping that future.
