Latest Headlines
Nigerian AI Researcher Exposes Pidgin Blind Spot in OpenAI, Google
Stories by Emma Okonji
The Meaning Intelligence Lab, an independent research initiative that was unveiled at the recent World Public Relations Forum (WPRF 2026) in Abuja by Dr. Celestine Achi, who is a Nigerian Artificial Intelligence (AI) Researcher, has exposed massive pidgin blind spot on OpenAI, Google and Anthropic Models, but fixed it in one paragraph.
The Meaning Intelligence Lab, which is the benchmark study of 500 Nigerian posts, showed frontier AI models scored as low as 11 per cent on Pidgin recognition compared to 92 per cent for human analysts—until a single paragraph of linguistic guidance lifted accuracy past 90 per cent.
He said the Meaning Intelligence Lab, took 500 real social-media posts, news excerpts and speeches in Standard English, Nigerian English, Nigerian Pidgin and code-mixed speech, adding that five Certified Meaning Intelligence Analysts labelled each item three times, independently, producing 1,500 expert judgements across nine dimensions of meaning, from surface sentiment and true intent to irony, coded subtext, risk and recommended communications action.
He explained that six AI models, at flagship and low-cost tiers, were then scored against exactly the same human reference.
Giving details of the findings of the study, Achi said AI misnamed how Nigerians speak.
According to him, “Models routinely labelled Nigerian Pidgin as ‘mixed English’ and Nigerian English as Standard English. The gap appeared for every vendor and every price tier.
“Knowledge of how Nigerians speak is the fix. When Meaning Intelligence Framework gave three models one paragraph of guidance on how Pidgin grammar works, their Pidgin recognition rose from 20–47 per cent to 90–96 per cent, level with the analysts,” adding that the team will confirm the result on fresh data in Sprint II.






