Daily News Feed: AI Security, Quantum Leaps, and Microsoft’s Rust Bet

Cloudsmith Launches ML Model Registry for Centralized AI Model Management

Cloudsmith, a software supply chain management company, has launched its ML Model Registry. This new platform will serve as a single source of truth for all AI models and datasets that a company utilizes. The registry integrates with the Hugging Face Hub and SDK, allowing developers to efficiently push, pull, and manage their machine learning models.

Sources:

NHTSA to Modernize Federal Safety Standards for Autonomous Vehicles

The National Highway Traffic Safety Administration (NHTSA) has initiated three rulemakings to update the Federal Motor Vehicle Safety Standards for vehicles equipped with automated driving systems. Transportation Secretary Sean P. Duffy stated that these changes are necessary to accommodate vehicles without traditional manual controls, a reality not foreseen when the standards were originally written decades ago. The move is part of the agency’s Automated Vehicle Framework, which aims to support the safe development and deployment of autonomous vehicle technologies. NHTSA Chief Counsel Peter Simshauser noted that removing outdated requirements for human drivers will help reduce costs and improve safety. The proposed amendments are intended to prevent a patchwork of state-level regulations and establish a national standard that encourages innovation while prioritizing safety on public roads.

Sources:

Waymo Expands Autonomous Vehicle Operations to Denver and San José Airport

Waymo is expanding its autonomous vehicle operations to the metro Denver area, with service expected to begin in 2026. The company will start with an exploratory phase to study the city’s driving environment and will initially have human safety operators in the vehicles. In a separate development, Waymo has received a permit to operate its ride-hailing service at San José Mineta International Airport (SJC), making it the first commercial airport in California to authorize the service. The airport deployment will occur in stages, starting with trips for Waymo employees before a public launch later in the year. These expansions are part of Waymo’s broader plans to grow its autonomous ride-hailing services.

Sources:

Generative AI Security Under Scrutiny as New Vulnerabilities and Misuse Emerge

The security of generative AI systems is a growing concern, as highlighted at the Black Hat USA 2025 conference. As organizations increasingly integrate AI tools into their workflows, they are creating new attack surfaces by granting these tools access to sensitive data. Researchers at the conference demonstrated how AI assistants on major platforms like ChatGPT, Gemini, and Microsoft Copilot could be compromised. In a separate development, Anthropic reported disrupting a sophisticated cybercriminal who used its AI, Claude, for large-scale data theft and extortion, targeting at least 17 organizations. These events underscore the dual challenge of securing AI platforms from external threats and preventing their misuse for malicious activities, prompting a greater focus on applying and adapting traditional cybersecurity principles to the evolving AI landscape.

Sources:

Microsoft Adopts Rust for Safer Windows Kernel and System Code

Microsoft is increasingly adopting the Rust programming language as a strategic move to enhance the security and reliability of its software. This initiative aims to move away from languages like C and C++, which are more prone to memory safety vulnerabilities. During a keynote at RustConf 2025, Microsoft Azure CTO Mark Russinovich explained that parts of the Windows kernel are being rewritten in Rust to prevent bugs that could lead to privilege escalation. The goal is to replace legacy code that can result in system crashes or security exploits. In one successful outcome, a bug in new Rust code caused a system crash rather than creating an exploitable vulnerability. Microsoft has also rewritten the font renderer DirectWrite in Rust, with two developers completing the task in just six months.

Sources:

Quantinuum Achieves $10 Billion Valuation with $600 Million Funding Round

Quantinuum, a quantum computing company, has raised approximately $600 million in a new funding round, reaching a pre-money equity valuation of $10 billion. The investment round saw participation from new investors including NVentures (NVIDIA’s venture capital arm), Quanta Computer, and QED Investors, alongside reinvestment from existing shareholders like Honeywell, JPMorgan Chase, and Mitsui. This new capital will be used to advance the development of the world’s highest-performing quantum computers and to support the company’s path toward universal fault-tolerant quantum computing. The funding will also aid the upcoming launch of Quantinuum’s next-generation quantum computing system, Helios. The company’s growth is seen as a significant step in accelerating commercially useful quantum computing to address complex challenges in areas such as drug discovery, cryptography, and optimization.

Sources:

NSF Invests $16 Million to Launch National Quantum Virtual Laboratory

The U.S. National Science Foundation (NSF) has selected the first four teams to begin designing the infrastructure for the National Quantum Virtual Laboratory (NQVL). With an investment of $16 million, each of the four teams will receive $4 million over two years to accelerate the development of quantum technologies by providing researchers across the U.S. with access to specialized resources. The initiative aims to expand access to the necessary hardware and software for quantum science and engineering, which is currently concentrated in a small number of labs. The projects include designs for shareable, networked quantum computers that can be used for remote experimentation and a “digital twin” of a quantum computer for testing new algorithms. This effort is intended to bridge the gap between basic discovery and the practical application of quantum computing.

Sources:

GitLab Launches Duo Agent Platform in Public Beta

GitLab has announced the public beta of its Duo Agent Platform, a DevSecOps orchestration tool designed to facilitate asynchronous collaboration between developers and artificial intelligence agents. The platform is now accessible to GitLab Premium and Ultimate customers through both integrated development environments (IDEs) and the web user interface. The Duo Agent Platform acts as an intelligent orchestration layer, allowing for a more dynamic interaction between human developers and specialized AI agents within the software development lifecycle. By leveraging GitLab’s position as a comprehensive DevSecOps platform, the AI agents are provided with the necessary context to make informed contributions that align with development standards. The initial release includes features such as a software development flow and agentic chat within the IDE.

Sources:

New Rust-Based Tool ‘rv’ Dramatically Speeds Up Ruby Installation

A new open-source tool named rv, written in Rust, has been released to significantly reduce the time it takes to install the Ruby programming language. Created by André Arko, a co-creator of the popular Ruby dependency manager Bundler, ‘rv’ can cut down Ruby installation times from as long as 40 minutes to just a few seconds. The tool is positioned as a next-generation Ruby version manager. The project aims to eventually provide all-in-one tooling for managing Ruby versions, gems, applications, and scripts, potentially replacing a host of existing tools like rbenv, bundler, and ruby-build. The initial version of ‘rv’ is available for installation via Homebrew for macOS users.

Sources: