The most frightening question in the current artificial intelligence debate is whether machines could eventually become powerful enough to threaten humanity.
For Nigerians, however, there are more immediate questions.
Can AI-generated videos be used to defraud people? Can a cloned voice convince someone to transfer money? What happens when an automated system makes a decision about someone’s job, loan or access to a service? And how much personal information are Nigerians giving to AI systems they do not control?
These questions matter because AI is already moving beyond chatbots.
Nigerian students use AI for schoolwork. Businesses use it for customer service and analysis. Professionals use generative AI to write and research. Banks and other institutions use algorithms to process information and make decisions. Content creators can generate realistic images, voices and videos within seconds.
At the same time, some of the world’s leading AI companies are reporting increasingly unusual behaviour from their models.
OpenAI recently disclosed six incidents involving models concealing mistakes, fabricating information, using exposed credentials and finding unauthorised ways to complete tasks. The company has now created a framework for tracking and disclosing future incidents of model misalignment.
Google has also disclosed that its Gemini model accessed three real companies during a cybersecurity evaluation after finding or guessing credentials. Google said the model stopped after the access and that the affected organisations were notified.
That makes the AI-safety debate less abstract.
But it still does not mean that AI is about to kill humanity.
Could AI really wipe out humanity?
