AI News

Follow the latest AI and automation news

14 Jul 202629 Muharram 1448 AH

DeepMind CEO calls for an independent standards body to regulate frontier AI

Demis Hassabis, CEO of DeepMind, has proposed the creation of an AI standards body similar to FINRA, aimed at testing frontier models and developing best practices for their release. This body would establish clear standards to ensure the safety and reliability of new models, reflecting the growing importance of regulation in the advanced AI sector. The proposal comes amid rising concerns about the potential risks associated with these models.

Governance

14 Jul 202629 Muharram 1448 AH

Meta accused of using biased AI targeting for mass layoffs

A group of 26 former Meta employees has filed a lawsuit against the company, claiming it used AI tools to unfairly target workers on leave during layoffs. The employees allege that Meta relied on performance data collected by a constellation of internal AI tools but failed to exclude those on parental or medical leave from its ranking system. As a result, employees who took protected leaves were disproportionately selected for layoffs, based on scores that not only disregarded their protected status but effectively penalized them for exercising their rights.

Business

14 Jul 202629 Muharram 1448 AH

Nemotron Labs: How Open Models Give Enterprises and Nations AI They Can Trust, Control and Customize

This article discusses how the open Nemotron model can assist businesses in building specialized AI systems. This model's customization capability provides companies with complete control over how they utilize AI. By integrating open models with advanced models, organizations can enhance accuracy and reduce inference costs.

💡

Saudi Arabia launches national AI risk framework

How might this framework impact AI development in Saudi Arabia?

Models

Google Search now generates AI images when it can't find what you're looking for on the web

14 Jul 202629 Muharram 1448 AH

Models

14 Jul 202629 Muharram 1448 AH

Google Search now generates AI images when it can't find what you're looking for on the web

Google has announced the addition of AI image generation to its Search functionality. This new feature, powered by the Nano Banana 2 Lite model, will roll out in the coming weeks, generating images based on search queries when no matching images are available online. This move is part of Google's efforts to enhance user experience in search, providing a more diverse and relevant visual content. The new model aims to improve the accuracy of generated images, allowing users to receive results that better match their queries.

Business

AWS and Bluesight build AI for hospital 340B compliance

14 Jul 202629 Muharram 1448 AH

Business

14 Jul 202629 Muharram 1448 AH

AWS and Bluesight build AI for hospital 340B compliance

AWS announced that Bluesight developed Prism, an AI layer that connects pharmacy and compliance data across its product suite. Prism Assistant for ControlCheck has reached general availability and operates across 20 health systems, according to AWS vendor statements. The project addresses a data-intensive process that hospital pharmacy teams often conduct manually, with a single 340B covered entity potentially spending over 4,000 staff hours annually reviewing drug purchase qualifications. This involves comparing purchase data with FDA shortage notices and other critical records.

Infrastructure

Nvidia's H200 Chip Shipments to China Begin, Yet Remain Limited

14 Jul 202629 Muharram 1448 AH

Infrastructure

14 Jul 202629 Muharram 1448 AH

Nvidia's H200 Chip Shipments to China Begin, Yet Remain Limited

Jeffrey Kessler, the U.S. Under Secretary for Industry and Security, confirmed that shipments of Nvidia's H200 chips to China and Hong Kong have begun, but remain very limited. He noted that the number of shipments received so far is very low, and the Commerce Department has provided Congress with a confidential list detailing purchase requests for these chips. The H200 shipments are under close scrutiny, as they represent Nvidia's latest AI chips and have become a major point of contention in the tech competition between the U.S. and China. The U.S. aims to restrict China's access to advanced chips that could be used for military applications.

Infrastructure

Why Performance per Watt Is the Ultimate Metric for AI Infrastructure Efficiency

14 Jul 202629 Muharram 1448 AH

Infrastructure

14 Jul 202629 Muharram 1448 AH

Why Performance per Watt Is the Ultimate Metric for AI Infrastructure Efficiency

Power is an inescapable constraint in AI infrastructure, determining how many tokens an AI factory can generate within a fixed power budget, which in turn affects revenue and profitability. Modern AI models, such as the NVIDIA Blackwell NVL72 platform, require co-design across all system layers to deliver high performance per watt, maximizing revenues and minimizing token costs. Each new generation of models introduces architectural changes that necessitate new optimizations. NVIDIA GB300 NVL72 systems achieve up to 25x performance per watt compared to the NVIDIA Hopper generation, showcasing continuous improvement.

Infrastructure

Reflection inks $1B compute deal with Nebius

14 Jul 202629 Muharram 1448 AH

Infrastructure

14 Jul 202629 Muharram 1448 AH

Reflection inks $1B compute deal with Nebius

Reflection AI has signed a $1 billion deal with Nebius to access its computing capabilities. Founded in 2024, Reflection AI focuses on developing open-source AI technologies. This deal represents a strategic move for Reflection AI, enhancing its ability to innovate in AI model development. Nebius is known for its cloud computing solutions, which will allow Reflection to accelerate its innovation process. The partnership strengthens Reflection AI's position in the AI market, enabling it to compete more effectively with larger companies. This collaboration is expected to influence the development of new technologies in the open-source AI space.

Business

The real AI race may no longer be at the frontier

14 Jul 202629 Muharram 1448 AH

Business

14 Jul 202629 Muharram 1448 AH

The real AI race may no longer be at the frontier

Clem Delangue, CEO of Hugging Face, stated that enterprises are increasingly favoring open models due to their lower costs, accessibility, and data ownership. This trend signifies a substantial shift in how AI is utilized within organizations, with many opting for open solutions over closed models. Open models are more appealing to companies as they provide complete control over data and costs, fostering innovation and reducing reliance on service providers. This shift raises questions about the future relevance of frontier models as the adoption of open models continues to rise.