Chinese AI startup Moonshot AI has launched Kimi K3, a 2.8 trillion-parameter AI model capable of tasks like coding, scientific research, and financial analysis. The model has gained recognition for its advanced AI capabilities and has completed complex engineering tasks autonomously.
However, due to overwhelming demand, Moonshot AI has paused new subscriptions to prioritize current users. This move comes as Chinese AI firms, including Alibaba with its Qwen 3.8 model, gain momentum and pose a challenge to American AI models like OpenAI's GPT-5.6 and Anthropic's Claude Fable 5.
In related news, Nvidia briefly lost its crown as the world's most valuable company to Apple, with the chipmaker's stock losing value as investors exercise caution around the AI boom. IBM's software sales have also been impacted by the surge in AI infrastructure spending, warning that revenue growth will be affected.
On the security front, a new survey aims to benchmark AI security decisions made by security leaders, while Capital One has open-sourced an internally developed AI-powered security tool called VulnHunter to find and fix software vulnerabilities at the code level.
Key Takeaways
['Moonshot AI launches Kimi K3, a 2.8 trillion-parameter AI model capable of tasks like coding and financial analysis.', 'Kimi K3 has gained recognition for its advanced AI capabilities, completing complex engineering tasks autonomously.', 'Due to high demand, Moonshot AI has paused new subscriptions for Kimi K3 to prioritize current users.', 'Alibaba has launched Qwen 3.8, a 2.4 trillion-parameter AI model that claims to be second only to Claude Fable 5.', "Nvidia briefly lost its crown as the world's most valuable company to Apple due to investor caution around the AI boom.", "IBM's software sales have been impacted by the surge in AI infrastructure spending.", 'A new survey aims to benchmark AI security decisions made by security leaders.', 'Capital One has open-sourced an AI-powered security tool called VulnHunter to find and fix software vulnerabilities.', "Kimi K3 is designed to compete with OpenAI's GPT-5.6 and Anthropic's Claude Fable 5.", 'Chinese AI firms are gaining momentum with the launch of Kimi K3 and Qwen 3.8, posing a challenge to American AI models.']China unveils Kimi K3, a trillion-parameter AI model
Chinese AI startup Moonshot AI has launched Kimi K3, a 2.8 trillion-parameter AI model that can handle tasks like coding, scientific research, and financial analysis. The model has gained widespread recognition for its advanced AI capabilities and has completed complex engineering tasks autonomously. However, due to overwhelming demand, Moonshot AI has paused new subscriptions to prioritize current users. Kimi K3 is designed to compete with OpenAI's GPT-5.6 and Anthropic's Claude Fable 5.
Alibaba unveils Qwen 3.8, a rival to Kimi K3
Alibaba has launched Qwen 3.8, a 2.4 trillion-parameter AI model that claims to be second only to Claude Fable 5. The model is designed to handle tasks like software engineering and long-horizon coding. Chinese AI firms are gaining momentum with the launch of Kimi K3 and Qwen 3.8, posing a challenge to American AI models.
Kimi K3 faces GPU crunch amid high demand
Chinese startup Moonshot AI has suspended new subscriptions for its Kimi K3 model due to high demand and limited computing power. The company is prioritizing existing subscribers and adding capacity to meet the demand. Kimi K3 is a 2.8 trillion-parameter large language model that has drawn global attention.
China's Kimi K3 model halts new subscriptions
China's Kimi K3 model has halted new subscriptions due to overwhelming demand. The model has gained popularity for its advanced AI capabilities, but the high demand has put a strain on the company's computing power. Existing subscribers are not affected, and the company is working to add more capacity.
Survey aims to benchmark AI security decisions
A new survey aims to benchmark AI security decisions made by security leaders. The survey, conducted by SANS Faculty Fellow Lenny Zeltser and Cyber Defense Matrix creator Sounil Yu, asks security leaders to share their experiences and decisions related to AI security. The goal is to create a benchmark that can be used by security leaders to make informed decisions.
Survey on AI security decisions now open
A survey on AI security decisions is now open. The survey, conducted by SANS Faculty Fellow Lenny Zeltser and Cyber Defense Matrix creator Sounil Yu, aims to benchmark AI security decisions made by security leaders. The survey takes only five minutes to complete and is open to anyone whose job includes securing AI.
IBM software sales hit by AI infrastructure spending
IBM's software sales have been impacted by the surge in AI infrastructure spending. The company has warned that its revenue growth will be affected due to the shift in corporate spending from software to data-center infrastructure.
Nvidia briefly loses crown as world's most valuable company
Nvidia briefly lost its crown as the world's most valuable company to Apple. The chipmaker's stock has been losing value as investors exercise caution around the AI boom.
Various loan products and services
Various loan products and services are being offered, including verification, servicing, next-gen, and flood cert products. The industry is also seeing an increase in AI and overall tech adoption.
AI product and service launches
D1g1t LLC launched the d1g1t MCP server, a connector built on the open model context protocol standard that links its enterprise wealth management platform directly to general-purpose artificial intelligence tools.
Charles Payne analyzes AI impact on economy
Charles Payne analyzes the impact of AI on the economy, discussing AI adoption rates, hyperscaler free cash flow forecasts, and semiconductor sector performance.
Harmonizing AI safety thresholds
A methodology has been developed to derive harmonized thresholds across three risk domains for frontier AI companies. The goal is to ensure consistent safety standards.
Training AI models
Training AI models may be the chance for a workplace power play. The knowledge required for training AI models is currently locked inside employees' heads.
Capital One open sources AI-powered VulnHunter
Capital One has released an internally developed AI-powered security tool called VulnHunter to the public as open source. The tool is designed to find and fix software vulnerabilities at the code level.
Sources
- What is Kimi K3? China's 2.8 trillion parameter AI model taking on Claude Fable 5 and GPT-5.6
- After Kimi K3, Alibaba unveils Qwen 3.8, claims it's second only to Claude Fable 5
- Moonshot AI’s Kimi K3 faces GPU crunch amid US-China tech rivalry
- China's new AI model halts new subscriptions as demand swamps capacity
- The AI Security Decisions No One Has Benchmarked Yet. A New Survey Fixes That.
- The AI Security Decisions No One Has Benchmarked Yet. A New Survey Fixes That.
- Spending on AI infrastructure hits IBM software sales
- Why Nvidia briefly lost its crown as the world's most valuable company
- Verification, Servicing, Next-Gen, Flood Cert Products; AI and Overall Tech Adoption
- AI Product & Service Launches
- Charles Payne: Hyperscaler investments continue to dominate the AI discussion
- Harmonizing AI Safety Thresholds
- Training AI models might be the chance for a workplace power play
- Capital One Open Sources AI-Powered ‘VulnHunter’ Security Tool
Comments
Please log in to post a comment.