2023 Elections: Technology expert explains how Atiku, Okowa, and Tambuwal’s voices were cloned

Dr Ndubuisi Ekekwe, a technology expert, based in Lagos, has said that the voices of Atiku Abubakar, Ifeanyi Okowa and Waziri Tambuwal in the leaked audio purportedly evidencing their plot to rig the February 25 presidential elections, were cloned.

He explained that cloning was achieved through artificial intelligence technology.

There had been earlier alerts that in the run-up to the 2023 Nigerian Presidential elections, a new kind of political attack campaign had emerged, that was using cutting-edge artificial intelligence technology to clone the voices of political figures.

The most recent example of this, according yo reports, involved the former Vice President Atiku Abubakar and his running mate Governor Ifeanyi Okowa, whose voices had been cloned using A.I. technology in an attempt to discourage voters before tomorrow’s elections .

The viral voice recording, which purports to be Atiku Abubakar making comments about INEC, has been circulating on social media in Nigeria, causing a stir among the electorate.

But technology experts had weighed in to offer explanations to doscount the genuinesness of the voices in the audio.

They pointed to irregularities that proved the audio tape was a fake, generated using sophisticated A.I. software that could mimic the sound and intonation of any human voice.

They cited online services like www.respeecher.com and www.Resemble.ai, which clone speeches that are indistinguishable from the original speakers’.

Dr Ekekwe particularly said: “Artificial intelligence is an incredible tool that has many useful applications, but it can also be used for nefarious purposes.

“In this case, someone has used A.I. to clone Atiku Abubakar’s voice and create a recording that never actually happened.

“We need to be vigilant and aware of these kinds of techniques, especially in the political sphere.

“I also hope the law enforcement authorities are getting more sophisticated in their investigations to hold those propagating these sort of dangerous activities”

He explained that “A.I. voice cloning technology has advanced rapidly in recent years, and it is now possible to create convincing fake recordings that are almost indistinguishable from real ones.

“However, there are ways to detect fake recordings, such as analyzing the waveform and frequency of the voice, which can reveal anomalies that are characteristic of A.I. generated speech.”