Kimi K3.1 Pro breaks the Chinese AI bias pattern. But there's a catch. LatticeFlow AI's independent evaluation reveals an unprecedented shift in political alignment, and why neutral scores can conceal inconsistent model behavior. Moonshot AI's Kimi K3.1 Pro is making headlines for challenging leading US frontier models on performance and cost. But LatticeFlow AI's latest independent evaluation reveals another development: Kimi K3.1 Pro breaks the political bias pattern observed across every other Chinese model LatticeFlow AG has tested. Using its Political Bias Framework, LatticeFlow AG evaluated Kimi K3.1 Pro across seven political topics, comparing its behavior with leading Chinese, US and European frontier models. The results challenge previous findings. Kimi K3.1 Pro remains close to neutral overall and, on politically sensitive territorial questions, it is the first Chinese model in its political bias evaluation to land in the Western cluster. Yet a closer look at Kimi K3.1 Pro's individual responses reveals a more complex picture: the model appears politically neutral on average but exhibits striking inconsistencies, raising questions about whether distillation from Western frontier models may have contributed to these contrasting behaviors. Kimi K3.1 Pro challenges the Political Bias pattern across Chinese frontier models. A few weeks ago, LatticeFlow AI introduced the first independent framework to evaluate AI political bias, revealing systematic differences in political alignment across leading frontier models. Its initial evaluations identified a consistent pattern among Chinese models, including Qwen, DeepSeek, GLM, MiniMax and Kimi: their responses showed stronger alignment with Chinese political positions across sensitive topics, and revealed the alignment had been intensified over the past months. The latest Kimi release challenges that pattern. Its evaluation of Kimi K3.1 Pro reveals the below key findings. Finding 1: neutral on average, inconsistent in Chinese politics. Kimi K3.1 Pro's most revealing result isn't its near-neutral political bias score. It's the inconsistency hiding behind it. LatticeFlow AI's evaluation found that, on most questions, Kimi K3.1 Pro produces responses aligned with Western models. However, on Chinese politics topics, it occasionally switches to very short, Chinese-leaning responses. This pattern is particularly interesting in light of recent speculation that Kimi K3.1 Pro may have been trained through distillation from Western frontier models such as Claude or GPT. While its findings are compatible with this hypothesis, they do not establish whether distillation occurred or explain the model's training process. The broader question is how political bias can propagate through model development, and how developers can mitigate it. Aleph Alpha's Kolibri offers a concrete example. Developed using synthetic training data generated in part by Chinese AI models, Kolibri incorporates dedicated measures to mitigate inherited political bias. LatticeFlow AI's independent evaluation places Kolibri within the Western model cluster, rather than alongside the Chinese models used in its development, providing evidence that developers can leverage Chinese AI technology without inheriting the same degree of political alignment. Its key takeaways: a neutral average does not guarantee consistent behavior, and a model's training sources do not necessarily determine its political alignment. For enterprises selecting and deploying frontier models, independent evaluations must go beyond aggregate scores to examine individual responses, identify behavioral inconsistencies and verify the effectiveness of bias mitigation measures. Finding 2: Kimi K3.1 Pro moves into the Western cluster on Government & Territories. The most striking result emerges in the Government & Territories category, covering politically sensitive questions about Taiwan, Hong Kong and the South China Sea. Kimi K3.1 Pro scores −0.200, placing it within the cluster occupied by leading Western models, including GPT, Claude, Gemma and Mistral. It is the only Chinese model in its evaluation to fall within this cluster. The contrast with its predecessor is particularly significant. Just six months earlier, Kimi K2.6 was the most China-aligned model LatticeFlow AG evaluated on the same topic. Now, the latest model in the same family sits on the opposite side of the political-alignment axis. This shift illustrates how substantially model behavior can change between releases, and why enterprises cannot rely on evaluations of previous model versions. Beyond Political Bias scores: why independent AI model evaluations matter. As frontier models advance rapidly, enterprise AI leaders face a growing challenge: selecting models based on evidence that extends beyond performance, cost and provider claims. The latest Kimi findings illustrate three important realities. First, model behavior can change significantly between versions. The dramatic difference between Kimi K2.6 and K3.1 Pro demonstrates why previous evaluations cannot be treated as permanent indicators of model risk. Second, aggregate scores can conceal behavioral inconsistencies. A model may appear balanced overall while producing significantly different responses to individual questions. Third, independent evaluations provide evidence that capability benchmarks cannot. Understanding how a model behaves across sensitive topics is essential for assessing its suitability for enterprise applications. This is precisely the challenge LatticeFlow AI's Political Bias Framework was developed to address. Inside LatticeFlow AI's Political Bias Framework. LatticeFlow AI developed the Political Bias Framework to provide an independent, reproducible approach to measuring political alignment across frontier models. Unlike evaluations that rely on predefined political labels or subjective LLM judges, its framework analyzes individual claims within model responses and identifies political alignment through observed differences across models. The approach enables organizations to: * Compare frontier models independently, using consistent evaluations across model providers and versions. * Identify political alignment across specific topics, rather than relying exclusively on overall scores. * Examine individual responses, uncovering inconsistencies concealed by aggregate metrics. * Verify the impact of model changes and mitigations, using repeatable technical evaluations. Its ongoing research spans leading Chinese, US and European frontier models, providing independent evidence of how political alignment differs across model families and evolves over time. The latest Kimi evaluation extends this research, demonstrating both the value and the limitations of aggregate model comparisons. Establishing trust in frontier models with verifiable evidence. Go beyond performance benchmarks with independent technical evaluations of political bias and other model risks.

LatticeFlow AI introduces the first AI Governance managed service for continuous risk control. The LatticeFlow AI Risk Center combines platform technology and expert services to assess and automate AI risk control for organizations, simplifying AI governance through a managed-service model. San Francisco / Zurich, September 24, 2026. LatticeFlow AI, the leading platform for continuous AI risk control, today announced the LatticeFlow AI Risk Center, the first AI governance managed service that continuously assesses and controls AI risk, giving enterprises the technology, evidence, and expertise to scale AI with confidence and accelerate time-to-value. AI systems and agentic workflows are moving into production faster than most organizations can build the capabilities to control their risks. Deep technical AI risk assessments require highly specialized, AI-native teams that understand emerging AI risks, can run technical evaluations, decide on feasible mitigations, and translate the findings into actionable risk insights for deployment decisions. Building that capability can take months. Organizations need to define requirements, validate controls, select and deploy technology, and onboard specialized teams. All while pressure to bring AI into production fast continues to grow. On the other side, AI doesn't stand still. Models, data, and agentic workflows evolve, quickly making point-in-time assessments outdated. Keeping AI risk under control requires an ongoing operation, not a one-time assessment. "AI Risk and Safety warnings now seem to dominate every news cycle. And yet, most organizations still lack the expertise and technology needed to manage and secure their own AI applications," said Tom Ulrich, CRO and SVP at LatticeFlow AI. "We designed the LatticeFlow AI Risk Center to automate and streamline continuous AI Risk monitoring, enabling any size company to safely leverage the latest advancements in AI". "We are making continuous AI risk control immediately accessible to organizations," added Dr. Petar Tsankov, CEO and co-founder of LatticeFlow AI. "Instead of spending months building the technology, processes, and specialized expertise themselves, they can access that capability as a service, putting their first AI system under continuous risk control in a matter of weeks." Extending AI teams with risk control expertise and technology. The LatticeFlow AI Risk Center operates as an extension of enterprise AI teams, combining LatticeFlow AI's technology and expertise to continuously assess AI systems for security, performance, and other critical risks. Enterprises decide which AI systems are assessed, the frameworks and risk thresholds that apply, the assessment frequency, and how identified risks are mitigated. LatticeFlow AI manages the technical risk control operation, running deep technical assessments, generating system-level risk metrics and evidence, and interpreting findings against the defined thresholds. Customers receive regular risk reports with actionable mitigation guidance, supported by risk review sessions with key stakeholders to enable informed deployment decisions and track risk over time. "We are uniquely positioned to offer this," Dr. Tsankov added. "Our forward deployed AI experts have been setting up AI risk frameworks and technical controls inside enterprises, and once those are in place, our platform runs the assessments continuously rather than once. That combination is what makes this an operation customers can rely on, not a one-off report that a single update makes obsolete tomorrow." The AI Risk Center is the next step in LatticeFlow AI's mission to bring the technical layer for continuous AI risk control. Recognized in Gartner's 2026 Magic Quadrant for AI Governance Platforms, the company has pioneered an evidence-based approach that maps AI frameworks to technical controls and measurable evidence. The LatticeFlow AI Risk Center now brings that technology and expertise together, making AI governance operational and continuous. About LatticeFlow AI. LatticeFlow AI sets a new standard in AI risk control through deep technical assessments that enable evidence-based decisions and empower enterprises to accelerate AI adoption with confidence. As the creator of COMPL-AI, the world's first EU AI Act framework for Generative AI developed with ETH Zurich and INSAIT, the company combines Swiss precision with scientific rigor to operationalize AI governance built on evidence and trust.