Terms

Large Language Model

Follow the latest coverage, related explainers and connected technology stories.

Latest coverage

CN
Japanese Clinic Adopts "Local AI" to Enhance Healthcare Efficiency

Japanese Clinic Adopts "Local AI" to Enhance Healthcare Efficiency

A doctor at the Hongo Masago Heart Clinic in Tokyo has successfully built a local AI system running on Mac Studio hardware, enabling the processing of sensitive patient data without sending it to the cloud, while achieving operational savings equivalent to an additional administrative staff member.

CN
Why Do Game Development Teams Need to Combine Static Analysis and Artificial Intelligence?

Why Do Game Development Teams Need to Combine Static Analysis and Artificial Intelligence?

JetBrains explains that artificial intelligence tools accelerate code writing, but they may increase the volume of problems and produce defects different from those caused by developers. The article proposes adopting static analysis as a deterministic and recurring foundation, while using artificial intelligence to address logical and more complex problems.

CN
How Can You Make AI Agents Predictable in Sensitive Systems?

How Can You Make AI Agents Predictable in Sensitive Systems?

The Stack Overflow Blog proposes confining the language model to a single nondeterministic node and leaving the rest of the system to ordinary, testable code. The methodology includes fixed flow diagrams, structured outputs, independent validation, human approvals, and an immutable log for every decision.

CN
Meta Uses AI to Detect Ads That Secretly Lead to Child Exploitation Material

Meta Uses AI to Detect Ads That Secretly Lead to Child Exploitation Material

Meta announced new artificial intelligence tools to detect ads and accounts that may be used to direct users to child sexual abuse material outside Facebook and Instagram. The company said it acted on 33.2 million items of this kind in the first half of 2026, and that its systems detected more than 97% of them before user reports.

CN
Homeward Raises $120 Million to Accelerate Home Buying and Selling in the U.S. Market

Homeward Raises $120 Million to Accelerate Home Buying and Selling in the U.S. Market

Homeward has raised a $120 million Series D round alongside securing a $330 million asset-backed debt facility. The company finances solutions that enable people to buy a new home before selling their current one or obtain a fast cash offer, while using artificial intelligence to reduce manual work in underwriting and transaction processing.

CN
Microsoft: Attackers Are Gaining an Early Lead Over Defenders in Leveraging AI

Microsoft: Attackers Are Gaining an Early Lead Over Defenders in Leveraging AI

In its 2026 Digital Defense Report, Microsoft warns that threat actors are currently benefiting from artificial intelligence faster than security teams, particularly in vulnerability discovery, malware development, and accelerating post-compromise activity. The company believes that the time needed to turn a discovered vulnerability into a weapon has fallen to less than 24 hours, although real-world campaigns still rely heavily on human direction.

CN
Moores Lab AI Bets on Agentic AI to Accelerate Chip Design

Moores Lab AI Bets on Agentic AI to Accelerate Chip Design

Moores Lab AI is developing tools to automate stages of chip design and verification, saying its customers have been able to reach their first bug within 48 hours instead of waiting months. But the company acknowledges that AI reliability remains a critical challenge in an industry that does not tolerate mistakes.

CN
MAVI Emerges from Stealth with $4 Million in Funding to Address the AI-Supported Accounting Skills Gap

MAVI Emerges from Stealth with $4 Million in Funding to Address the AI-Supported Accounting Skills Gap

MAVI, an AI-powered hiring marketplace connecting U.S. companies with global finance and accounting talent, announced its emergence from stealth and revealed a $4 million seed funding round. The platform has more than 3,000 accounting professionals and focuses on recruiting experienced professionals who can use AI tools and review their outputs.

CN
U.S. Court of Appeals Rejects Expanding Copyright Liability in AI Model Training Case

U.S. Court of Appeals Rejects Expanding Copyright Liability in AI Model Training Case

The U.S. Court of Appeals for the Ninth Circuit rejected using Section 1202 of the Digital Millennium Copyright Act to create a new basis for liability based on the absence of copyright information from the outputs of a new model. The decision ends a legal theory advanced by anonymous GitHub contributors against OpenAI and Microsoft, while other contractual claims in the case remain.

CN
Materials Engineering Moves from Optimizing Results to Designing Questions with Artificial Intelligence

Materials Engineering Moves from Optimizing Results to Designing Questions with Artificial Intelligence

The article explains how artificial intelligence, particularly large language models and agents, is changing the materials development cycle, from selecting experiments to designing the questions that guide research. It argues that human oversight and the design of verification loops remain an essential part of any reliable system.

CN
FDA Updates List of AI-Enabled Medical Devices and Prepares to Track Foundation Models

FDA Updates List of AI-Enabled Medical Devices and Prepares to Track Foundation Models

The U.S. Food and Drug Administration has updated its list of AI-enabled medical devices authorized for marketing in the United States, emphasizing that the list is not exhaustive and does not represent a complete database of all devices. The agency is exploring future mechanisms to identify devices that incorporate foundation models, including large language models and multimodal systems.

CN
OmniOps and HPE Sign Memorandum of Understanding to Develop Sovereign AI Solutions in Saudi Arabia

OmniOps and HPE Sign Memorandum of Understanding to Develop Sovereign AI Solutions in Saudi Arabia

OmniOps and HPE signed a memorandum of understanding during LEAP to develop AI solutions that combine HPE-produced infrastructure in Saudi Arabia with locally developed OmniOps platforms and in-Kingdom deployment capabilities. The partnership aims to help government entities and companies move AI use cases from experimentation to operation, while keeping the infrastructure, software and management processes within Saudi jurisdiction.

CN
Turkey Launches Program to Train 550 Researchers in Strategic Technologies During Its First Year

Turkey Launches Program to Train 550 Researchers in Strategic Technologies During Its First Year

A new Turkish program aims to train 550 doctoral students in artificial intelligence, cybersecurity, information technologies, and computer systems during its first year, with the number set to reach around 3,000 researchers within five years. Participants will take up research assistant positions while pursuing their studies at 23 research universities.

CN
Ramp Launches Router to Route AI Requests Between Multiple Models

Ramp Launches Router to Route AI Requests Between Multiple Models

Ramp has launched Router, a service that allows companies to use multiple large language models and switch between them through an API, with tools for selecting a model based on cost, performance, or request difficulty. The service is currently available only in the United States and is free through the end of 2026, excluding inference costs.

CN
Türkiye Puts Its 2026–2030 Artificial Intelligence Plan into Effect with Targets Including $10 Billion in Investments

Türkiye Puts Its 2026–2030 Artificial Intelligence Plan into Effect with Targets Including $10 Billion in Investments

Türkiye’s Artificial Intelligence Plan for 2026–2030 took effect after its publication in the Official Gazette. It includes targets to train 5 million people, increase data center capacity to 1 gigawatt, and stimulate more than $10 billion in investments.

CN
CNCF Grants Kubeflow Graduated Project Status to Strengthen the Standard for Running AI on Kubernetes

CNCF Grants Kubeflow Graduated Project Status to Strengthen the Standard for Running AI on Kubernetes

The CNCF announced Kubeflow’s graduation, deeming it a mature platform ready for production use in managing the artificial intelligence and machine learning lifecycle on Kubernetes. The decision follows the project’s expansion to more than 6,600 contributors and nearly 260 million downloads of its Python packages from PyPI.