Four journalists across Canada, India, Germany and the USA reveal what it is really like to train the AI models that are reshaping — and possibly destroying — their own careers.
By Imran Malik | Media & Technology Desk | MediaBites.com.pk
Courtesy: Reuters Institute for the Study of Journalism
Trained reporters and editors across four countries are using their journalism skills to teach artificial intelligence systems how to write, fact-check and edit — work that could eventually eliminate the jobs they hoped to hold.
They are doing it because they need the money.
Journalists Find AI Work Where Newsroom Jobs Are Not
Darius Osborne did everything right. He wrote for his university newspaper at Howard University in Washington, D.C., interned in newsrooms and won scholarships before graduating, ready to start his journalism career.
What he got instead was a job at Meta, labelling AI-generated content for quality and accuracy.
“The process of applying for journalism jobs is completely horrible,” Osborne said. “I was scrambling after college and just applying, applying, applying.”
In Canada, Khaleda Khan spent four to five months after graduating searching for entry-level media work before joining xAI — Elon Musk’s artificial intelligence company — as a trainer for the Grok chatbot.
For six months, she corrected transcription data, trained Grok to understand Hindi and Urdu, reviewed AI-generated videos for errors and identified hallucinations on high-stakes topics, including politics, crime and breaking news.
“There were a lot of transferable journalism skills I was really able to use,” Khan said. “There were not really as many people in tech as there were people in the humanities. There were a lot of writers there.”
In India, freelance journalist Bisma Farooq joined Invisible Technologies for AI training and data annotation work after five years of freelancing still was not covering her bills. The work triggered an immediate identity crisis.
“Once I entered this, I had an existential crisis: What am I doing? What is this?” Farooq said. “I eventually kept working, but it did not sit well with me.”
In Germany, Bettina Blass — a freelance journalist since 2003 — joined AI startup tisix.io after noticing a sharp decline in writing assignments she attributed directly to AI. For 18 months, she wrote rules and prompts teaching an AI tool to convert press releases and police alerts into news articles.
“Journalism is always written in the form of an inverted pyramid. As a journalist, you know that, but the AI tool does not,” Blass said. “You need to know the basics of journalism in order to teach an AI tool how to write that way.”
Workers Question Whether They Are Training Their Replacements
The same question followed all four journalists through their AI work.
Khan discovered her xAI contract would not be renewed after six months. “While I was working there, I was thinking, am I helping make something that will replace me?” she said. “I didn’t know it was going to happen so quickly.”
Osborne said he often thinks he is training his own replacement, but reached a conclusion that others in the industry are reluctant to say plainly.
“If this whole thing were a vehicle, AI would be the car, and humans would be the gas,” Osborne said. “We are definitely the fuel behind the whole thing.”
His daily work inside AI systems convinced him that the technology is far less independent than its public image suggests. Significant human labour still powers every AI system consumers use. The machine needs the journalist far more than it advertises.
All Four Say AI Cannot Replicate the Human Core of Journalism
Despite their different experiences, all four journalists arrived at the same conclusion: AI cannot replace what is most essential about reporting.
“After going through this AI training, I feel that the human touch — the real emotions and real feelings of a real journalist — will never be replaced by a machine,” Farooq said.
Blass and Khan argued AI is well-suited to routine tasks — sports scores, press releases, traffic updates — which could free human reporters to focus on higher-impact work. In their view, AI reshapes journalism collaboratively rather than purely destructively.
Khan offered the sharpest assessment of where journalism now stands.
“Journalism is already in this weird transitional phase, where not many people are reading, but people do want content, and journalism is not the same as content,” she said.
Pakistani Journalists Face the Same Pressure With Fewer Safeguards
For Pakistani journalists and media organisations, the pattern described across four countries is not a distant warning.
Pakistan’s journalism market — already under severe economic pressure, with shrinking advertising revenue and political constraints — has virtually no institutional framework to navigate AI’s disruption. Pakistani freelancers are already quietly taking AI training work from international platforms to supplement incomes that journalism alone cannot sustain.
The question for Pakistan’s media industry is whether its institutions will engage with this transition strategically — or simply watch it happen.
