Open Source
Microsoft Introduces Agent Control Specification for AI
TechCrunch AIhigh confidence
Why it matters
- →ACS provides a unified framework for managing AI agent behavior, enhancing security and compliance.
- →It simplifies the auditing process by integrating controls into a common governance layer.
- →Developers can now ensure consistent agent behavior across different environments and frameworks.
Microsoft has launched the Agent Control Specification (ACS), an open-source standard designed to help developers manage AI agent behavior more effectively. ACS allows developers to define policies that dictate what an AI agent can and cannot do, ensuring consistent behavior across different environments. This specification aims to streamline control mechanisms, making them easier to audit and apply across various frameworks. By integrating these controls into a common governance layer, Microsoft seeks to enhance security and compliance in AI applications.
Read originalMore from TechCrunch AI
© TechCrunch AIMarket & Regulationbusiness
Lovable expands Google Cloud usage fivefold
Lovable, a rapidly growing Stockholm-based startup, has signed a multiyear deal with Google Cloud to significantly increase its usage, reportedly by five times. This expansion includes enhanced access to Anthropic's Claude and Google's Gemini AI models, positioning Lovable to leverage advanced AI capabilities for its coding tasks. The deal also integrates Lovable's offerings into Google's enterprise agent marketplace, potentially simplifying procurement for enterprise clients. This collaboration not only boosts Lovable's growth prospects but also aligns with Google's strategy to support its substantial capital expenditure plans.
TechCrunch AI
© TechCrunch AIInvestment · $85B
Market & Regulationbusiness
Alphabet raises $85B for AI expansion
Alphabet's $85 billion stock sale marks a significant moment for AI investment, reflecting a strong market interest in AI-related ventures. The company initially aimed to raise $40 billion, but due to overwhelming demand, it secured $45 billion in the first tranche, with Berkshire Hathaway contributing $10 billion. This capital is set to enhance AI infrastructure and data centers, aligning with Alphabet's strategic focus on AI growth. The success of this offering indicates a favorable climate for upcoming AI IPOs, such as Anthropic and OpenAI. This development highlights the substantial financial backing AI projects are receiving from public markets, suggesting a robust future for AI investments.
TechCrunch AI
© TechCrunch AIVideo & Creative AIwriting
Google Launches Dreambeans AI App for Lifestyle Suggestions
Google Labs has introduced Dreambeans, an AI-driven app designed to animate users' lives by generating personalized lifestyle suggestions. By leveraging data from Google services like Gmail, Calendar, and Photos, Dreambeans curates daily stories that inspire users with new ideas and activities. The app aims to combat doomscrolling by limiting the number of stories to 10-14 per day, encouraging users to engage with the world beyond their screens. Available to U.S.-based Google AI Ultra subscribers, Dreambeans represents a novel approach to integrating AI into everyday life, offering a unique blend of personal intelligence and creativity.
TechCrunch AI
More in Open Source
Open Sourcemodels
llama.cpp b9489 release enhances CUDA support
The b9489 release of llama.cpp brings notable improvements for CUDA users, specifically by reserving space for quantized key-value caches at startup. This update also addresses previous feedback and removes certain assertions in the ggml-cuda.cu file, enhancing the CUDA experience. While it doesn't introduce new models or quantization techniques, the release continues to refine the platform's compatibility across macOS, Linux, and Windows. With ROCm 7.2 and KleidiAI support, llama.cpp is becoming a more robust tool for developers working with CUDA and other environments. This iteration is a step towards making llama.cpp a more versatile and efficient tool for AI development.
llama.cpp Releases
Open Sourcemodels
llama.cpp b9490 Release Expands Platform Support
The latest b9490 release of llama.cpp continues its trend of broadening platform compatibility, though with some notable exceptions. While macOS Apple Silicon users see KleidiAI support disabled, the release strengthens its Linux offerings with Vulkan and ROCm 7.2 support on Ubuntu. Windows users benefit from CUDA 12 and 13 DLLs, enhancing GPU performance options. Despite some features being disabled, this update demonstrates llama.cpp's commitment to being a versatile inference runtime across diverse systems.
llama.cpp Releases
Open Sourcemodels
llama.cpp b9493 Release Expands Platform Support
The b9493 release of llama.cpp continues to broaden its platform reach, notably integrating ROCm 7.2 for Ubuntu x64, which offers better support for AMD GPU users. Although features like KleidiAI on macOS Apple Silicon remain inactive, the update emphasizes extending functionality across various systems, including Vulkan support for both Ubuntu and Windows. While no new models are introduced, this release strengthens llama.cpp's role as a versatile inference runtime across multiple operating systems. Developers can now take advantage of improved GPU support, making it a more inclusive tool for those working outside the NVIDIA ecosystem.
llama.cpp Releases