By Ronnie Wabomba
NATIONAL
Media practitioners from various media houses in Uganda have been urged to use Artificial Intelligence (AI) responsibly without sacrificing the core principles of journalism.
Speaking during a Media League webinar held earlier this week on Zoom, organized by Consent Uganda with support from the Uganda Communications Commission (UCC), facilitator Eng. Dr. Peter Byamugisha from Deep Shift AI, said AI presents both opportunities and risks. He said AI should be viewed by journalists as a co-pilot, not an autopilot.
Byamugisha explained that media practitioners can use AI for transcription, translation, faster research, data analysis, idea generation, and improving the quality of their writing. However, he cautioned that AI also poses risks, including fabrication of information, hidden bias, privacy breaches, overreliance, copyright violations, and erosion of public trust.
He noted that one of the biggest misconceptions about AI is that many people believe it is always accurate and objective. Others assume AI understands information the way humans do, that if it sounds confident it must be correct, or that everything typed into AI tools remains private.
Byamugisha warned that AI reflects the data on which it is trained and can therefore be biased, outdated, or simply wrong. In his explanation, AI predicts likely words based on patterns rather than genuine understanding or knowledge of the truth. He added that AI can “hallucinate” by convincingly inventing facts, quotations, and sources, cautioning that many AI tools may learn from users’ inputs unless users explicitly opt out.
