How Generative AI Erased 40,000 Academic Writing Jobs in Kenya

Once a lucrative lifeline for university graduates in Nairobi, Kenya’s shadow-scholar economy vanished in under two years.

For over a decade, Mr. Bandi ran a thriving digital business out of Nairobi. Holding a degree in public health, Bandi chose to write academic essays, research papers, and dissertations for university students in the United States, the United Kingdom, and Australia rather than enter the local job market. At his peak, he charged between $40 and $70 per assignment, completing over 2,500 papers and generating earnings several times higher than entry-level local salaries.

Today, Bandi’s essay-writing business no longer exists.

His story mirrors a silent economic collapse across Kenya. An estimated 40,000 highly educated Kenyan youth who earned a living as “shadow scholars” have seen their livelihoods erased—not over decades, but in less than 24 months following the release of OpenAI’s ChatGPT and subsequent large language models.

The Rise of the “Shadow Scholars”

Kenya’s academic writing boom was born out of a stark economic paradox: high digital literacy and widespread English fluency alongside an official youth unemployment rate exceeding 25%.

With over 100,000 university graduates entering the Kenyan job market each year into an economy dominated by informal labor, digital freelancing became a vital economic release valve. Supported by national initiatives like the Ajira Digital Program, thousands of graduates turned to global freelance platforms and direct agency networks.

          [ Western College Students ]
                        │  (outsourced assignments)
                        ▼
           [ Global Freelance Brokers ]
                        │  (subcontracted work)
                        ▼
      [ ~40,000 Kenyan University Graduates ]
      (Earning $900 – $1,200/mo writing essays)

By the early 2020s, Nairobi, Eldoret, and Kisumu had become world hubs for academic ghostwriting. Full-time writers routinely earned between $900 and $1,200 per month—a middle-class income in a country where the average monthly wage hovered under $300.

The Two-Year Industry Collapse

When ChatGPT launched in late 2022, the impact was immediate. Western college students—the end clients of this supply chain—realized they no longer needed to pay $50 for a custom-written essay when a generative AI model could produce a passing draft in seconds for free.

According to research from the Center for AI Safety and Scale AI, the share of end-to-end freelance tasks that AI models can perform autonomously leapt from 2.5% to over 16% in less than a year. The “text-in, text-out” commodity layer of knowledge work was the first to fall.

“I never imagined AI could do this kind of work. A.I. is coming for bankers, for accountants, it’s coming for engineers, and A.I. will come for architects. It’s coming for everybody.”

— Teresios Bundi, former academic writer, to The New York Times

From Essay Writers to AI Trainers

For forced survivors of the collapse, higher-level knowledge jobs have not automatically appeared. Instead, workers have shifted sideways into lower-paying piecework—ironically helping build the tools that automated their previous careers.

• Data Annotation & Moderation: Former writers like Alex Munya moved into content moderation and data labeling for global AI training platforms, where earnings are significantly lower and less predictable.
• Photo & Media Editing: Agency owners like Richie Esilabalaba, who once managed over 100 ghostwriters, shuttered their writing operations to focus on basic image processing and digital assistant tasks.
• “Text Humanizing”: A remaining niche involves taking raw AI-generated essays from students and re-editing them—adjusting cadence, introducing mild stylistic variance, and removing predictable syntax patterns to bypass anti-plagiarism software like Turnitin and GPTZero.

A Warning Signal for the Global Workforce

While academic ghostwriting carried ethical controversy as “contract cheating,” the speed at which the industry vanished offers a crucial case study for global labor economists.

Unlike theoretical projections about future automation, Kenya’s academic writing sector presents clear before-and-after data. It demonstrates that when a profession’s primary deliverable is standard text or code delivered remotely, artificial intelligence can replace an entire task layer almost overnight, leaving workers without a transition buffer.

As governments in developing nations continue to push digital skills initiatives, Kenya’s experience highlights a critical lesson: digital labor strategies built purely on cheap, remote execution are highly vulnerable to rapid AI automation. Sustainable economic growth will require transitioning workers up the value chain toward tasks requiring judgment, accountability, and complex verification