‎SYPALA: AI Users must verify outputs before trusting them – John Sitsofe Mensah

‎Technology Associate at IMANI Center for Policy and Education and Technology policy analyst, John Sitsofe Mensah, has urged users of artificial intelligence (AI) to verify information generated by AI systems before relying on or publishing it.
‎
‎Speaking at the Students and Young Professionals African Liberty Academy (SYPALA) 2026, on August 26, where he led a discussion on “AI and the Evolution of Research: Live, hands-on Labs on Secure AI Prompt
‎Engineer and Verification Loops.”

Mensah noted, users must not treat an AI-generated response as truth, regardless of how detailed the instructions given to the system are.
‎
‎“Once it comes, don’t take that first one as truth,” he said, stressing the need for users to remain actively involved in reviewing AI-generated information.
‎
‎According to him, AI should serve as an analytical partner rather than replace human reasoning.
‎
‎“AI should be our analytical Partner the co-pilot, right? It means you are The pilot. Remain that pilot. Do not ever outsource your reasoning to what? The computer,” he said.
‎
‎He recommended that users ask AI to fact check and self audit its outputs, trace information back to its sources and refine the work where inaccuracies are identified.
‎
‎“Your reputation is tied to that. If you’re known for putting out wrong information, you’re done. Nobody will trust you,” he cautioned.
‎
‎Mensah, who is also a Technical Program Manager at Google, explained that AI systems can hallucinate and fabricate information, making verification essential, particularly when the information will be used for research, publication or decision making.
‎