Elon Musk's AI posted a series of far-right messages. Musk has previously attempted, unsuccessfully, to force his political views onto Grok.
Details. On July 4, Musk claimed that xAI had “improved Grok significantly” and said users should have noticed a difference in its responses. This followed an earlier post where he asked for training data that is “politically incorrect” but “factually true.”
► The AI soon began responding to posts with antisemitic conspiracies, praising Hitler as a defender of “white identity,” and echoing other far-right talking points on feminism and the Civil Rights Movement.
► In May, Grok had promoted Musk’s views by pushing the “white genocide” conspiracy theory in South Africa, even inserting it into unrelated queries. In some cases, the AI contradicted itself in the same post.
► After backlash in both cases, the posts were removed. In the case of the “white genocide” responses, xAI blamed an “unauthorised modification.” Regarding the more recent content, Musk said the chatbot had become “too compliant to user prompts.”
Context. AI’s capacity to generate tailored, mass-produced messaging makes it a valuable tool for the capitalist class, which seeks to shape public opinion as users increasingly rely on AI for search, news, and decision-making.
► Earlier AI models were criticised for having a “left-wing liberal bias” by politicians across the US political spectrum. In response, companies like OpenAI, Meta, and Google adjusted their outputs to appear “neutral” and avoid politicised content.
► AI has already been accused of influencing US political processes. During the 2020 election cycle, concerns were raised about algorithmic bias, targeted misinformation, and AI-generated content that amplified partisan narratives.
Important to Know. Musk’s attempts reveal one way AI is to be used by capitalists. They are likely to adopt similar AI tools to propagate their views with subtler messaging and tighter controls to avoid Musk’s PR failures.
► Just days after Musk’s far-right failure with Grok, xAI secured a $200 million contract with the Pentagon, further deepening the integration of AI into the US state.
► This is part of a broader trend. In the US, executives from major AI firms sit on the Pentagon and Army advisory boards. In the UK, the AI Frontier Taskforce includes both state and corporate leadership. The NHS is increasingly relying on private AI systems for diagnostics and data management.
► This trend reflects the growing fusion of state power and tech monopolies. Although AI technology itself is just a tool, when used by the capitalist class, it serves as an instrument for ideological control, reinforcing dominant narratives and protecting class rule.