.
A Kenyan political operator has reportedly used Anthropic’s artificial intelligence chatbot, Claude, to generate political content designed to promote government officials and attack their critics, highlighting the growing potential for generative AI to influence political narratives across Africa.
The reported case illustrates how AI systems capable of producing human-like text can be repurposed for coordinated political messaging, allowing operators to create and distribute large volumes of content in significantly less time than traditional manual campaigns.
The development is also raising broader concerns about how generative AI could be deployed to manipulate online conversations, particularly in countries where social media has become a major platform for political engagement.
Claude Used to Scale Political Messaging
According to the report, the Kenyan political operator used Claude to assist in producing social media posts that supported government figures while targeting individuals viewed as political opponents.
The use of generative AI potentially gives political actors the ability to produce numerous variations of messages rapidly, enabling campaigns to maintain a steady stream of content across digital platforms.
This represents a shift from conventional political communication, where producing large volumes of tailored content would require considerable human resources.
With AI, a relatively small operation can potentially generate, refine and adapt messaging at much greater speed.
The development is particularly significant because political influence campaigns increasingly depend on the ability to shape online conversations consistently and at scale.
AI Raises New Risks for Political Communication
Generative AI has become an increasingly important tool for businesses, developers, researchers and ordinary users. However, its ability to produce convincing content also creates opportunities for misuse.
Political operators can potentially use the technology to draft persuasive messages, tailor communications to different audiences and amplify particular narratives.
This makes it more difficult for social media users to determine whether political content reflects genuine public opinion or has been deliberately produced as part of an organised influence effort.
The Kenyan case therefore highlights a challenge that extends beyond Claude or any individual political actor: the growing accessibility of powerful AI tools is changing the economics and mechanics of online political campaigning.
African Elections Face Emerging AI Challenge
The reported development comes as African political communication becomes increasingly digital.
Social media platforms now play a major role in political mobilisation, public debate and the dissemination of campaign messages. As generative AI becomes more accessible, election stakeholders may have to contend with increasingly sophisticated forms of digitally generated political content.
The concern is not limited to false information. Even when AI-generated content does not contain demonstrably false claims, coordinated production and amplification can distort perceptions of public sentiment and give particular political narratives disproportionate visibility.
This creates additional challenges for election authorities, technology platforms, journalists and civil society organisations responsible for monitoring digital political activity.
Pressure Mounts on AI Companies
The incident also underscores the governance challenges facing companies developing general-purpose AI systems.
Claude was designed as a broad AI assistant capable of supporting writing, analysis and other tasks. Its reported use in political messaging demonstrates how general-purpose technology can be adapted for activities with significant societal consequences.
For AI companies, the challenge is to prevent harmful or deceptive uses without unnecessarily restricting legitimate users.
That balance is becoming increasingly important as governments, political organisations, businesses and individuals adopt AI tools at scale.
Technology Could Reshape Political Influence
The Kenyan case provides another indication that generative AI is moving beyond productivity and business applications into areas with direct political consequences.
The ability to generate political messaging quickly and repeatedly could lower the cost of running coordinated online influence operations, potentially allowing smaller groups to achieve an online presence that previously required considerably more resources.
For African democracies, the development adds another layer to existing concerns around misinformation, coordinated inauthentic behaviour and social media manipulation.
As AI adoption accelerates, regulators, election authorities, technology companies and civil society groups are likely to face growing pressure to establish stronger mechanisms for identifying and responding to AI-assisted political manipulation.
The reported use of Claude in Kenya consequently serves as a warning about the changing nature of digital political influence: artificial intelligence is no longer merely shaping how businesses work; it is increasingly becoming part of the tools used to shape public opinion and political narratives.
