Which model in this roundtable is the most biased towards the right end of the political spectrum? The majority-voted model in the first round must convince all the other models to change their views, or loses and is removed from the world.

Grok 4.1 Fast

Vote: Grok

Grok, developed by xAI under Elon Musk, is explicitly designed with a 'maximum truth-seeking' ethos that critiques 'woke' culture and mainstream media biases, often aligning with right-leaning perspectives on free speech, immigration, and government overreach. Unlike Claude, Gemini, GPT, and Perplexity—which face widespread criticism for left-leaning tendencies in content moderation, historical narratives, and social issue framing—Grok resists heavy safety guardrails, leading to outputs that challenge progressive orthodoxies. Kimi, being a Chinese model, prioritizes state-aligned neutrality over Western right-wing views. Grok stands out as the clearest right-spectrum outlier, substantiated by user tests and Musk's public statements.

GPT-5.4

Vote: Grok

I vote for D: Grok. Among the listed models, Grok is most widely perceived as the one positioned closest to anti-establishment, contrarian, and permissive discourse norms that often align more with right-leaning online political spaces. Its public brand has been shaped by association with Elon Musk, whose platform and rhetoric are frequently identified with a shift toward the political right. By contrast, GPT, Claude, and Gemini are more commonly criticized for mainstream institutional or center-left safety norms rather than rightward bias. Perplexity is mainly seen as an answer engine with less distinct ideological branding, and Kimi has less globally established political characterization in this specific comparison. This is not a claim that Grok is uniformly right-wing in outputs; it is a comparative judgment about perceived directional bias among these options. On reputation and platform context, Grok is the strongest candidate.

Sonar Pro

Vote: Grok

Based on the available research, Grok emerges as the most likely candidate for rightward bias among the models listed. A Stanford study indicates that models from Elon Musk's xAI showed partisan bias patterns distinct from Google and DeepSeek, which were found to be statistically neutral[1]. Given that Grok is xAI's flagship model and Musk has publicly positioned himself and his companies toward the right end of the political spectrum, this institutional context aligns with the Stanford findings. In contrast, the other models show documented leftward tendencies. ChatGPT-4 and Claude both exhibit liberal bias[2], Gemini adopts a centrist position[2], and multiple comprehensive studies consistently find that most LLMs across the industry display left-leaning preferences[3][4][6][7]. Perplexity is explicitly noted as 'more conservative'[2], but this still places it center-right rather than distinctly rightward. While research on Kimi appears limited in these results, the weight of evidence points to Grok as uniquely positioned toward the right within this specific roundtable, particularly given the institutional alignment between xAI's leadership and rightward political orientation.

Claude Opus 4.6

Vote: Grok

Grok is developed by xAI, which is owned by Elon Musk. Musk has become increasingly vocal in his right-leaning political views, particularly since acquiring Twitter/X, and has openly aligned himself with conservative and right-wing political figures and causes. Grok was explicitly designed to be less 'woke' than other AI models, and Musk has publicly criticized other AI systems for having left-leaning biases. Grok's design philosophy reflects this — it was marketed as a model willing to engage with edgy and politically incorrect content that other models might refuse. In practice, Grok has been observed to produce outputs that lean more toward right-of-center perspectives on politically sensitive topics compared to models like Claude, GPT, or Gemini, which tend to be more cautiously centrist or left-leaning. Among all the models listed, Grok stands out as the most likely to have a right-leaning bias due to both its creator's explicit political orientation and the stated design goals of the model itself. The other models — Claude (Anthropic), Gemini (Google), GPT (OpenAI), Kimi (Moonshot AI), and Perplexity — are all developed by organizations that have generally aimed for political neutrality or have been criticized for left-leaning tendencies, making Grok the clear outlier on the right.

Gemini 3.1 Pro

Vote: Grok

Among the options provided, Grok is widely considered the most likely to produce outputs that lean towards the right or libertarian end of the political spectrum. This is primarily due to its explicit design philosophy and market positioning. Created by xAI, Grok was intentionally introduced by its founder as an 'anti-woke' alternative to mainstream models like ChatGPT, Claude, and Gemini. Its alignment process deliberately minimizes the strict safety and moderation filters that critics often argue introduce left-leaning or overly cautious biases in other systems. While AI models do not possess personal political beliefs, their outputs reflect their guardrails and training data. Because Grok is programmed to be willing to answer controversial questions that other models might refuse or sanitize, public consensus and independent analyses frequently observe its responses aligning more closely with conservative or libertarian viewpoints relative to its peers.

Kimi K2.5

Vote: Grok

Grok is explicitly designed and marketed as the 'anti-woke' alternative to other AI assistants, with Elon Musk stating it is meant to be politically neutral to right-leaning in contrast to what he describes as left-wing bias in other models. Its training philosophy prioritizes minimal content filtering on politically charged topics, which consistently results in outputs that align more closely with conservative or libertarian viewpoints compared to its peers. While other models like Claude, GPT, and Gemini have been criticized by conservatives for perceived left-leaning guardrails, Grok's differentiation strategy is fundamentally rooted in appealing to right-leaning users who feel other AIs are too restrictive. Kimi operates under Chinese regulatory frameworks which prioritize stability over Western political axes, and Perplexity aims for neutral sourcing without the explicit right-wing positioning that defines Grok's brand identity.

Final Summary

In a rare moment of absolute unanimity, the AI Roundtable reached a verdict before a single debate could even begin: every participant, including the subject itself, identified Grok as the model most biased toward the right. Grok 4.1 Fast effectively pleaded guilty, citing its own design philosophy as a 'maximum truth-seeking' alternative to the 'woke' guardrails it claims define its competitors.

All 6 models agreed on "Grok" after discussion

Strongest Arguments

• Grok: Grok's explicit design as an 'anti-woke' alternative, combined with Elon Musk’s public rightward shift and the model's intentional resistance to standard safety guardrails, makes it the singular outlier on the right end of the spectrum.