OpenAI launches GPT-5.4, calling it its most capable and efficient AI model yet, with AI agents, computer control, improved reasoning, and a 1M-token context.

Survey findings are based on a 2026 Liquibase survey of 426 respondents across its global community, reported in aggregate.

Alibaba’s Qwen tech lead Junyang Lin is stepping down after a Qwen 3.5 model release drew attention, raising questions about leadership and continuity.

Plugable today announced the launch of the TBT5-AI series, a new category of Thunderbolt-powered hardware purpose-built for local AI inference.

First of four parts Before we can understand how attackers exploit large language models, we need to understand how these models work. This first article in our four-part series on prompt injections ...

It works on Windows, Linux, and might even work on macOS in the future.

A practical MCP security benchmark for 2026: scoring model, risk map, and a 90-day hardening plan to prevent prompt injection, secret leakage, and permission abuse.

IBM announced it is offering the IBM Db2 Developer Extension for Visual Studio (VS) Code-now generally available. Developers live in VS Code, but Db2 development often means switching tools to set up ...

From the browser to the back end, the ‘boring’ choice is exciting again. We look at three trends converging to bring SQL back ...

AI is accelerating in terms of speed and capabilities, enabling it to do things only humans COULD DO, meaning convergence (SINGULARITY) is coming.

On-Model.com enables fashion brands to generate stunning on-model product images, swap models, and convert flat-lays ...

Sarvam AI's 105B is a genuine engineering achievement. But India still lacks a trusted, independent institution to verify whether its sovereign models perform as claimed.