Ground Truth.
AI, checked against the source.
← 2026-11-172026-11-18later →

OpenAI’s GPT-6.2 models improve some benchmarks while coding results split

2026-11-18

OpenAI released GPT-6.2 Astra and Sol with lower Astra API rates and mixed independent coding results.

models · openai · coding

OpenAI’s GPT-6.2 safety tests expose a gap between AI monitors

2026-11-18

OpenAI’s system card reports sharply different detection results across monitors in simulated GPT-6.2 security tests.

cybersecurity · ai-security · red-teaming · agent-safety

Microsoft rolls out a portfolio of specialized MAI models

2026-11-18

Microsoft announced a family of specialized coding, security, image, audio and decision models across its existing products.

microsoft · models · products

ONNX Runtime’s GitHub repository became unavailable while packages stayed online

2026-11-18

The canonical ONNX Runtime GitHub repository returned 404 responses while existing packages remained listed in major registries.

open-source · infrastructure · supply-chain

Coding agents sped up pull requests as review delays grew in a large firm study

2026-11-18

A study of 655 firms linked coding-agent adoption to more pull requests and longer elapsed review times.

coding · agents · research · productivity

A search experiment finds fewer publisher clicks in forced AI Mode

2026-11-18

A field experiment found that forced Google AI Mode reduced click-through to publishers under its study definition.

search · publishers · research

Nvidia’s revenue geography does not show where its chips end up

2026-11-18

Nvidia’s latest revenue table classifies sales by direct-customer headquarters, not final chip destination.

nvidia · chips · finance

Tencent’s Hy4 Flash puts a large checkpoint behind a conditional license

2026-11-18

Tencent’s Hy4 Flash offers downloadable weights, but its full-precision repository is about 440.5 GB and its license has commercial conditions.

models · open-weights · tencent

← 2026-11-172026-11-18later →