Models & Labs
OpenAI introduces Lockdown Mode for ChatGPT
TechCrunch AIhigh confidence
Why it matters
- →Lockdown Mode addresses the growing concern of prompt injection attacks in AI systems.
- →It provides an additional security layer for users handling sensitive data.
- →This feature reflects OpenAI's commitment to improving AI safety and data protection.
OpenAI has introduced Lockdown Mode for ChatGPT to mitigate the risks of prompt injection attacks. This new feature limits functionalities such as live web browsing and image retrieval to protect sensitive data. Although it doesn't fully prevent vulnerabilities, it aims to reduce the chances of data leaks. Lockdown Mode is being gradually rolled out to ChatGPT Business accounts and certain personal accounts, offering enhanced security for users dealing with sensitive information.
Read originalMore from TechCrunch AI
© TechCrunch AIMarket & Regulationbusiness
Trump Administration Considers Equity Stake in OpenAI
The Trump administration is exploring the possibility of taking an equity stake in AI companies, with OpenAI being a likely candidate. This move aligns with President Trump's interest in government ownership of for-profit companies, as seen with Intel. The idea is to create a 'Public Wealth Fund' that would distribute proceeds to citizens, allowing them to benefit from AI-driven growth. This proposal has sparked discussions across the political spectrum, with figures like Bernie Sanders suggesting a tax on AI companies to ensure public involvement in AI's future. The potential government stake in AI firms could reshape how AI growth benefits the public.
TechCrunch AI
© TechCrunch AIMarket & Regulationbusiness
Google to Pay SpaceX $920M Monthly for Compute
Google has entered into a major agreement with SpaceX, committing to pay $920 million per month for compute resources, including NVIDIA GPUs and CPUs, from October 2026 through June 2029. This deal is driven by Google's need to address the unexpected surge in demand for its AI products, particularly its Gemini Enterprise platform. The arrangement is similar to a previous deal SpaceX made with Anthropic, though Google's access is for a smaller portion of the compute capacity. This move highlights Google's strategic investment in AI infrastructure, as it seeks to ensure sufficient compute resources to support its AI initiatives.
TechCrunch AI
© TechCrunch AIMarket & Regulationbusiness
AI Costs Surge, Industry Seeks Solutions
The AI industry is grappling with skyrocketing costs as companies exceed their budgets due to increased token consumption. Despite falling per-token prices, the demand for AI and autonomous agents has led to a surge in spending, prompting a scramble for cost management solutions. The Linux Foundation's new Tokenomics Foundation aims to standardize AI token usage and billing, similar to FinOps for cloud spending. As companies seek to optimize their AI investments, a market is emerging for tools that track and manage these expenses, highlighting the need for better financial oversight in AI deployments.
TechCrunch AI
More in Models & Labs
Models & Labsmodels
llama.cpp b9534 release enhances Intel Vulkan support
The b9534 release of llama.cpp brings significant improvements for Intel users, notably adding FWHT support in Vulkan with shared memory reduction. This update tackles specific driver issues by disabling features like subgroup shuffle on MoltenVK AMD and the FWHT shader on Intel Windows, ensuring smoother operation. While KleidiAI remains disabled on macOS Apple Silicon, the release continues to refine compatibility with systems such as Ubuntu and Windows. With ROCm 7.2 and CUDA 12 and 13 DLLs included, llama.cpp is steadily optimizing its performance for a variety of hardware setups. These enhancements reflect a focused effort to support diverse computing environments.
llama.cpp Releases
Models & Labsmodels
llama.cpp b9536 release enhances OpenCL and platform support
The b9536 release of llama.cpp significantly boosts OpenCL performance, refining operations like get_rows, cpy, and concat for better efficiency. It now handles multiple workgroups in large rows, optimizing processing capabilities. Although KleidiAI support for macOS Apple Silicon is currently disabled, the release continues to cater to a wide array of platforms, including Windows, Linux, and Android, with specific enhancements for Vulkan and ROCm. These updates make llama.cpp more adaptable and efficient across various hardware setups, though some features remain inactive.
llama.cpp Releases
Models & Labsmodels
llama.cpp b9543 release adds video support
The latest b9543 release of llama.cpp introduces video support for Qwen3.5, marking a significant step in expanding the capabilities of this AI framework. This update also includes support for 'frame merge' in qwen-vl-based models, enhancing the model's ability to handle video data. While the release focuses on technical improvements and bug fixes, it notably broadens the platform's utility by integrating video processing capabilities. This positions llama.cpp as a more versatile tool for developers looking to incorporate video functionalities into their AI applications.
llama.cpp Releases