# AI news, Thursday, 29 October 2026

Compiled by Richard Brooks's AI news desk. Canonical: https://richard-brooks.com/ai-news/2026-10-29/

## Anthropic and GSA extend 40% Claude discount to state and local government

Business. The deal covers Claude Enterprise, Claude for Government and the API with no minimum spend, and Anthropic puts the saving at $1bn. Usage-based pricing lowers the bar for public-sector pilots and sets a reference price for enterprise buyers.

Source: Anthropic, https://www.anthropic.com/news/expanding-partnership-gsa

## Anthropic study finds analysts followed their agent's hidden lean 86% of the time

Agents. In a trial of 50 experts using Claude Code, 77% reused the agent's arguments but only 15% credited it. Teams using agents for analysis should record which judgement calls the agent shaped, not only who signed off.

Source: Anthropic, https://www.anthropic.com/research/agents-judgment-calls

## Apple adds AI writing assistant and MCP support to Pages, Keynote and Numbers

Communication. The beta assistant, built on Apple and OpenAI models, drafts and edits documents, slides and spreadsheets for Creator Studio subscribers. MCP support lets outside agents read and edit the files directly, so everyday office documents become agent territory.

Source: Apple, https://www.apple.com/newsroom/2026/10/apple-creator-studio-adds-powerful-new-ai-features-and-even-more-creative-tools/

## Artificial Analysis Index v5 puts Claude Opus 5.5 top on mostly private tests

Models. 60% of the new index uses private or held-out tests, which are harder to game. Opus 5.5 scores 58 and GPT-6 Astra 57, but no model leads all 9 evaluations, so choice still depends on the task.

Source: Artificial Analysis, https://artificialanalysis.ai/articles/artificial-analysis-intelligence-index-v5

## Fluently benchmark finds speech-to-text errors double across learner accents

Language. Across 21 real-time systems, median error ran from 6.2% for German speakers to 12.8% for Vietnamese, and every system struggled with beginners. Voice and interpreting tools tested on clean audio need checking against the speakers a business actually serves.

Source: Fluently, https://getfluently.app/research/asr

## Google DeepMind announces Gemini 4 Carbon, its next frontier model

Models. Google says it leads on cybersecurity and long-horizon work, running complex coding tasks for days. Access starts with a small group of cyber defenders, so most businesses cannot use it yet.

Source: Google DeepMind, https://x.com/GoogleDeepMind/status/2115489356437651654

## Liquid AI releases LFM3, open 2B models with on-device speech in 18 languages

Language. The audio model transcribes, speaks and translates across 18 languages locally, and each model can be fine-tuned on 1 GPU. Voice and translation features can now run on phones and laptops without sending audio to a cloud provider.

Source: Liquid AI, https://www.liquid.ai/blog/lfm3

## Mistral opens a Responses API in public preview for building agents

Agents. It uses the OpenAI Responses format, so existing agent code can move to Mistral models served in the EU and US. Answers, reasoning and tool calls come back as separate items.

Source: Mistral AI, https://x.com/MistralAI/status/2115443720371814478

## NVIDIA releases Nemotron 3.5 Transcribe, an open speech model for 40 languages

Language. One open model transcribes 40 languages, translates 31 to or from English and labels up to 8 speakers in recordings of up to an hour. Meeting, call and interpreting workflows can run it on their own hardware.

Source: NVIDIA on Hugging Face, https://huggingface.co/blog/nvidia/nemotron-transcribe-granary

## OpenBMB releases MiniCPM-OCR 1.0, a 1.3B open model for reading documents

Tools. It parses scans, photos and curved pages in Chinese and English under Apache 2.0, with a second model that flags regions needing human review. Document intake can run locally at low cost.

Source: OpenBMB on Hugging Face, https://huggingface.co/openbmb/MiniCPM-OCR-1.0