The Disruption of Music Production and Culture in Kenya Through Artificial Intelligence

By Samuel Mwangi Ngari

Today the music industry has gone through many transformations one of them being Artificial Intelligence (AI) which raises optimism among Kenyan artists but equally raises fears and concerns at the same time.

How does AI tools generate music? AI tools use machine learning to make analysis of already existing enormous data from language models to give similar output that is easily understood by human beings. Similarly in music, AI tools use machine learning algorithm to analyze an artist(s) catalogue in efforts to create a similar composition based on the analysis.

AI tools through deep learning try to mimic the human’s brain structure to become more flawless compared to humans, it uses artificial neuron networks with multiple layers to learn and make decisions as described by the Media Council’s, Media Guide on The Use of Artificial Intelligence in Kenya.

Many African artists have joined other global artists in using AI tools in music production since it boomed in November. Artificial intelligence like preceding technologies is a double-edged sword but AI proves to be more unpredictable, since its technology continues to evolve and even its innovators admit its unpredictability.

In 2023, American Grammy winning band ‘The Beatles’ released their record ‘Now and Then’, which used AI tools and techniques to resurrect a demo by the late John Lennon. The, first AI-assisted song to ever in a Grammy, according to Loudwire, marked a monumental moment for the band’s fans since their last release which was 45 years ago after their split. The song has more than 74million views on YouTube streaming platform.

The demo had been named not usable but using AI technology to isolate vocals through de-mixing, they restored and generated John Lennon’s voice to release their record which went mainstream reuniting fans across the world and it also reflects the need for artists to be flexible with changing trends.

AI is causing a lot of positive and disruptive change in the music industry. In Kenya artists and non-artists have joined the bandwagon of AI users to create music. In Africa AI tools are making music production easier, lowering its original cost and one might argue removing barriers for new upcoming artists and making music production accessible to them. This will give new artists recognition in an already flooded market but this is just the tip of the iceberg.

Upcoming artists, veteran producers and upcoming producers can edit videos, transcribe data, generate images, generate lyrics and connect ideas easily. AI tools enhance human creativity, helping songwriters with writers’ block finish their songwriting process faster and helping producers refine records.

AI is combating the bureaucracy that has been originally associated with the music industry, enabling artists plan on release dates, release music without necessarily needing music labels. While AI tools are easing the penetration for artists in the music industry, some fear these tools may replace human creators making most jobless.

In Kenya this would mean increasing unemployment levels while people continue protesting low wages and unfair pay from telecommunication companies; by doing so AI might have done more harm than good to people’s livelihoods.

AI tools while creative, their output lack to match human output by sounding cliché and boring like Mc Cann, an American artist comments on the use of AI. For AI generative tools to make a similar song to an artist it can take up to nine hours for the original or sampling artists to be satisfied, on the other hand professional artists like Nicki Minaj and the likes of Lady Gaga have admitted to writing song within 10 minutes.

Many professional artists today still prefer to write their own songs, since they get inspired easier by a beat made by a human creator or AI tools. AI lacks emotions, feelings and basic human experiences making human output to be far much better in quality. AI continues to pose a threat to human creators by devaluing creative work, hence flooding streaming serving with low quality outputs preventing quality artists to standout.

Ironically older pro-artists continue to reign int the music industry, for instance the late Angela Chibalonza and her peers have constantly reigned in the Gospel industry from the 2000s regardless of modern music coming out meaning sometimes quality supersedes quantity. However, some fear if these AI outputs especially low-quality outputs, are left unregulated many may lead to overloads blocking authentic work from getting recognition.

AI tools are being used by artists to promote their music using AI algorithms which target audiences of a certain genre and new fans too. However, these algorithms are being used to breach privacy rights and to manipulate listeners without giving them much freedom to choose their preferred taste. Many audiences are unable to identify AI produced content, enabling manipulation. Audiences who face less freedom to choose their own music preferences may start experiencing psychological reluctance and reject new records and artists who may appear influenced by AI.

To solve this AI produced content should be well labelled for instance, some platforms like Instragram are currently labeling AI generated content, using Community Notes like employed by X (formerly Twitter) to give audiences the freedom to choose. Detection skills should also be imparted to audiences and tools to detect AI generated content be made readily available to audiences.

With evolving technology AI tools are likely to discover new music genres and blend them with existing ones, however this can lead to promotion of cultural appropriation if not used correctly. AI tools will also have a huge impact on diversity, especially for certain genres. Most AI tools have inadequate data on African culture, sounds and data.

This can lead to biased outputs which are probably analyzed from stereotypes already existing in the internet leading to skewed representations. AI tools also can lead to the homogenization of certain genres ignoring others that might be equally African, for instance, Afrobeats do not represent the whole of Africa but mostly Nigeria exempting other less known genres like Mugithii.

Responsible AI ensures fairness, transparency, accountability and considering the impact on society, privacy and the environment throughout the AI lifecycle. AI is not immune to flaws and it is not perfect.

The World Intellectual Property Day 2025 was celebrated around the world on 26th of April. In Kenya , the Copyright Act (Cap 130), 2001 protects various works by Kenyan artists from copyright infringement. The Act makes provision for copyright in literary, musical and artistic works, audio-visual works, sound recording and broadcasts.

AI practitioners should take into consideration authorship and ownership of generated content and works created by AI tools. However, in today’s world many AI users especially in music refuse to follow these rules.

Companies like META arguing publicly that data used to train models is unequivocally fair use since the data used is already public, makes it hard for countries like Kenya to pass Copyright laws. Recording labels who sue these AI tool companies to later have behind the scene agreements to make profits, also make it even harder to give original artists proper attribution.

AI tools have greatly made it easier for people to spread piracy works and making it hard for creatives to have control over their own voices in music. Many African countries having no clear rules regulating the use of AI and weak implementations which may cause harm to the music industry. In Africa where these piracy works are very high and unregulated lead to musicians being compensated unfairly, erasure of African legacies and cultural appropriation.

To combat this AI tools can be regulated and improved to allow original creators get compensation for every work sampled by new creators and AI tools. KorinAL in Nigeria, to avoid copyright issues associated with scrapping and flagging online data, the company employees African vocalists to create its datasets. Other AI tools like Suno can follow suit.

Who is benefitting most with the high overdependence of AI tools by artists, producers, music labels and the audience? There is a popular notion that if its free, you are the product especially social media and other technologies. Overdependence of AI tools by creatives hinder creativity and this negatively affects the music industry with creativity being its core.

Lastly AI affects music distribution and streaming of music content nowadays. Artists today rely on digital sales and digital streams to earn a living unlike in the past where physical album copies, CDS and vinyls were popular. While physical album copies and vinyls still remain popular, digital sales are more popular today.

With the rise of bot streams and fraud streams artists have greatly been impacted and further continue to suffer removal of streams suspected to be fraud streams. Since early April 2025, Spotify has removed streams it considers “artificial” at an unprecedented scale.

Artists have reported, sudden drops in streaming numbers and to make matters worse no way to appeal or clarify the data. Nigerian artist Davido reportedly lost over 18 million streams due to a recent purge by Spotify streaming platform without a clear justification (According to Hot97). While some may argue these moves are unfair, it further illustrates how AI continues to damage the music industry.

The first recording studio was established in Nairobi, particularly in the 1930s and 95 years down the line, the music industry in Kenya continues to change in unimaginable heights. Artificial Intelligence in music production, if exploited in the right ways we are going to reap the good fruits of technology. This will only be achieved by regulating the use of AI in this country and the rest of the world to address already existing ethical issues and upcoming ones since AI continues to evolve.