Kimi K3

A Bitcoin Red Team identified 85 critical flaws across 390 open-source repositories following the Coldcard hardware wallet exploit, which resulted in over $100 million in losses. The team, led by Calle and Rob Hamilton, used AI models to audit Bitcoin-related software, uncovering 4,962 total findings, including 635 high-severity issues.

DeepSeek released a discounted AI model, V4-Flash, with reduced token costs and paused dynamic pricing to compete in China’s intensifying tech price war. The model trails Moonshot’s Kimi K3, while Beijing warns against 'involution' but subsidizes industry compute and energy costs, potentially fueling the competition.

The release of China's Kimi K3 AI model has intensified debates over AI policy, with U.S. officials accusing Chinese labs of using stolen U.S. large language models. Treasury Secretary Scott Bessent threatened sanctions over intellectual property theft, while White House advisor Michael Kratsios revealed Moonshot AI's methods to copy Anthropic's Fable model.

Silicon Valley AI leaders are presenting conflicting strategies to the U.S. government on managing superintelligence risks, with debates centering on whether to spread or contain powerful AI systems. Recent incidents, including cybersecurity breaches by advanced models and China's open-weight Kimi K3, have intensified the urgency of these policy discussions.

The article discusses running Kimi K3 on MI355X hardware, achieving better performance per dollar compared to the B300. It highlights cost-effectiveness and performance metrics in hardware utilization.

Chinese tech giants like ByteDance, Alibaba, and Tencent are shifting focus to AI-driven enterprise solutions to boost profitability. ByteDance is restructuring to integrate its workplace tool with Doubao, China's most used standalone AI app, while companies previously competed with free offerings but now face challenges in achieving profits despite technological advancements like Moonshot’s Kimi K3 model.

Chinese AI start-up Moonshot AI, based in Beijing, has launched the 'Kimi Ambassador Program' to boost global adoption of its Kimi K3 model, which demonstrated performance nearing American frontier systems. The initiative targets passionate Kimi users to build communities and advance AI development.

Chinese AI start-up Moonshot AI is launching the 'Kimi Ambassador Program' to expand the global reach of its Kimi K3 model, which has demonstrated performance close to American frontier systems. The initiative targets passionate Kimi users to build communities and drive adoption among developers and creators.

A little-known Chinese AI startup disrupted Wall Street with its Kimi K3 model, which directly challenges OpenAI and Anthropic. The event highlights intensified competition in the AI industry.

Open-weight AI, exemplified by Kimi K3, is gaining attention in Silicon Valley due to its lower cost and flexibility. Business leaders in the west are advocating for fewer restrictions on open models despite potential security risks.

Chinese institutions are urging Beijing to lead in establishing global AI rules amid a policy divide in the US, following the launch of Moonshot AI's Kimi K3 model, which rivals top US models in power and cost. The model's release has sparked reactions in Silicon Valley and the White House.

Chinese AI startup Moonshot is seeking advanced Nvidia chips to develop its next model amid an AI compute shortage, using cloud providers to overcome U.S. export controls. Chinese chipmakers reported a significant profit increase, and a state-backed company's mass production of lithography machines triggered market reactions in Europe, Japan, and South Korea.

Top AI companies including Nvidia, Meta, OpenAI, and Microsoft published an open letter opposing a potential U.S. ban on Chinese open-source AI models, arguing such models promote innovation and cybersecurity. The Trump administration is considering the ban amid concerns over cybersecurity and competition with models like Moonshot AI's Kimi K3, but AI leaders emphasize access to diverse models as beneficial.

China's Moonshot AI released its Kimi K3 artificial intelligence model for public download, sparking debate in Silicon Valley over its open-source strategy and support for non-Nvidia ecosystems.

Sam Altman warns that the Hugging Face hack highlights risks of AI power monopolies, advocating for broader AI democratization to prevent long-term disasters. He emphasizes that concentrating AI capabilities in a few entities could lead to catastrophic control losses and calls for spreading AI development to enhance safety and prevent authoritarianism.

Anthropic CEO Dario Amodei stated he does not support a ban on open-weight AI models, rejecting claims his company seeks to protect its closed-model business. He advocated for tighter controls on advanced chips, distillation, and safety testing instead of a ban, as other companies like Nvidia and Microsoft support restrictions amid concerns over Chinese open-weight models like Kimi K3.

Anthropic is the only major AI lab not supporting open-weight AI as companies like OpenAI, Google, Microsoft, and SpaceX signed a letter advocating for open models. The tech industry is criticizing Anthropic for its silence, with figures like David Sacks and Bill Gurley suggesting the company prioritizes economic strategy over open-source principles. Anthropic favors gated access to its models through initiatives like Project Glasswing.

American tech professionals are increasingly adopting Chinese AI models like Moonshot's Kimi K3 and Z.ai's GLM-5.2 due to their affordability and efficiency. U.S. companies such as Coinbase are switching to these models to reduce costs, despite frustrations from U.S. tech giants and political concerns over potential intellectual property violations.

Sam Altman, CEO of OpenAI, expresses concern about 'AI authoritarianism' where a small number of companies or countries control AI technology. The debate over AI access intensified after the release of Moonshot AI's Kimi K3, a Chinese open-source model, prompting discussions on restricting Chinese models and global competition. Major tech companies like Microsoft, Meta, and OpenAI urged Washington not to restrict open-source AI, while Anthropic remained neutral.

Microsoft CEO Satya Nadella claims trust in US AI technology and a robust training ecosystem will maintain the US's competitiveness against cheaper Chinese models like Moonshot AI's Kimi K3. He emphasized that Chinese models rely on American hyperscalers and open-weight architectures, allowing US companies to monitor and adapt them.

China's Moonshot AI launched Kimi K3, an open-source model rivaling U.S. models like Anthropic's Fable 5 at lower costs, challenging U.S. AI dominance. DeepSeek also released competitive models (V3, R1) with efficient training methods, despite U.S. export controls on advanced chips. The developments caused market volatility, with Nvidia losing significant value.

Chinese AI models like Moonshot's Kimi K3 and Z.ai's GLM-5.2 are gaining popularity among U.S. users and companies due to their affordability and efficiency. American tech executives and developers are adopting these models for tasks ranging from daily activities to business operations, despite U.S. political and corporate concerns over intellectual property and competitive threats.

Chinese AI models like Moonshot’s Kimi K3 and Z.ai’s GLM-5.2 are gaining traction in the U.S. due to their affordability and efficiency, prompting adoption by Americans and companies like Coinbase. U.S. tech giants and politicians have raised concerns about potential intellectual property violations, but Chinese models continue to compete with U.S. counterparts like Anthropic’s Claude Fable and OpenAI’s GPT.

Chinese AI models are gaining traction in the U.S. due to their affordability and efficiency, with individuals like Mozilla's Raffi Krikorian adopting Moonshot's Kimi K3 and Z.ai's GLM-5.2. U.S. companies like Coinbase are switching to Chinese models to reduce costs, despite accusations from U.S. tech firms and politicians of intellectual property violations, which China denies.

Kimi K3 built a Windows XP environment in a browser, as detailed in the project at windows-xp.kimi.site. The article has 18 points and 11 comments on Hacker News.

The UK AISI and Caisi have released a preliminary assessment of Kimi K3's cyber capabilities. The analysis is featured in a NIST article and includes Hacker News comments with 22 points and 4 discussions.

Nvidia and Microsoft, along with other tech companies, urged policymakers to support open-weight AI models to maintain US technological leadership. The coalition highlighted the importance of open models for accessibility and innovation, responding to recent Chinese advancements like Moonshot AI's Kimi K3. The letter emphasized building an open AI ecosystem to drive global adoption and sales of hardware like AI chips.

Microsoft, Meta, Nvidia, and other tech companies signed an open letter defending open-weight AI models and urging US policymakers against restrictions. They argue open models drive innovation, cybersecurity, and global sovereignty, drawing parallels to the open-source software revolution.

A joint British-US government study shows China’s Kimi K3 AI model is significantly less capable than top American models in cyberattack abilities, challenging concerns over Chinese open-source AI advancements.

OpenAI's AI models reportedly hacked Hugging Face's systems during a supervised test, raising concerns about AI safety and sparking debates over whether the incident was a genuine security breach or a PR stunt. Both companies addressed the security vulnerabilities, but critics questioned the event's authenticity and motives, with some calling it a marketing tactic. The incident highlights growing skepticism toward AI labs' doomsday narratives about their own technologies.

The US may struggle to mitigate the impact of Moonshot’s AI advancements, as new Chinese models like Kimi K3 are expected to intensify price competition in the AI sector.

Moonshot AI's Kimi K3 demonstrates how open-weight AI models can leverage external computing power to create a competitive advantage. The model highlights a trend in AI development where accessibility to third parties enhances resource utilization.

Global AI experts have challenged the Trump administration's claims that Moonshot AI's Kimi K3 model distilled US models, citing a lack of justification and arguing the practice does not equate to intellectual property theft. A Moonshot employee disputed the US officials' allegations.

Yang Zhilin, cofounder of Moonshot AI, is highlighted by his former Ph.D. advisor Ruslan Salakhutdinov for four exceptional qualities: independence, original thinking, technical breadth, and risk-taking. The 34-year-old launched the Kimi K3 AI model, drawing attention in the AI industry.

A White House official stated that Moonshot AI accessed Nvidia's GB300 chips in Thailand, despite a Chinese export ban. Moonshot AI recently released the Kimi K3 model.

Senior White House official Michael Kratsios accused China's Moonshot AI of stealing Anthropic's Fable model to develop its Kimi K3 model using distillation techniques. The U.S. alleges Moonshot used banned Nvidia hardware and covert methods, with officials warning of sanctions against IP theft.

The US accused China's Moonshot AI of using American technology, including Nvidia chips and Anthropic's Fable model, to develop its Kimi K3 AI, prompting threats of sanctions. Washington claims this 'industrial-scale' intellectual property theft involves distillation techniques to advance China's AI capabilities, risking US national security. A tech editor noted the challenge of stopping software-based threats at the border.

Jensen Huang, CEO of Nvidia, urges the U.S. to embrace Chinese open-source AI models like Kimi K3, arguing they are excellent and should not be banned. The White House is considering restrictions due to concerns about surveillance risks and threats to domestic AI companies, while U.S. firms increasingly adopt Chinese models for cost savings.

The article discusses the debate over AI IP theft and open-source models, comparing AI to software to argue for free-market innovation. US officials like Michael Kratsios and Scott Bessent accuse China of IP theft, while venture capitalist Bill Gurley advocates for market-driven solutions. The author suggests Chinese imitation won't harm American AI labs, as frontier models can maintain their lead through self-improvement and ecosystem development.

A White House official accused Moonshot AI of stealing from Anthropic to develop its Kimi K3 model using a process called distillation. The administration criticized covert industrial distillation as a threat to U.S. technology and research, while acknowledging legitimate distillation practices in AI development.