🚨 AI News | TestingCatalog

Claude turns into Super App, adding web browser for Cowork

Recent Claude Desktop builds point to a built-in browser for Cowork sessions and a user-facing model comparison mode. The unreleased features suggest Anthropic is positioning Claude Desktop as a broader all-in-one AI workspace.

🗞 #anthropic @testingcatalog

OpenAI is updating its privacy policy in regard to ads on ChatGPT, mentioning ad personalisation.

> Ads are expanding to more regions and still only shown to free and Go accounts.

> Users will have to explicitly opt it into ad personalisation if they want to.

> In some regions, Free users can still choose not to see ads but they will be limited in a number of messages.

Google is working on a way to let Gemini Notebook query files from Google Drive directly from the chat UI.

> You can now query Google Drive files directly in Gemini Notebook. Ask a question to get started.

Gemini Notebook as a super app?

ANTHROPIC 🔥: Claude may get a new Model Comparison interface for users to have a simple way to see the difference in responses from different models.

> Users will be able to turn memory on and off, as well as "hold" responses.
> Unlike in an A/B testing UI, it looks like users will be able to switch between normal chats and model comparison mode themselves on demand.

Built-in LM Arena 👀

ANTHROPIC 🔥: Claude desktop will get a dedicated Browser inside that app that will always be accessible in Cowork sessions! A new Super App is rising.

This looks familiar 👀

ICYM 👀: Gemini users can now disable visible watermarks for images, videos, and songs. The setting isn't available in countries where they're legally required.

SymnthID would still remain there 🤖

OPENAI 👀: ChatGPT got a bunch of updates recently. Users can now automatically generate quizzes from the chat, search for reservations, add Google Drive files to the ChatGPT Library, and get better suggestions on the home page (paid plans only).

Anthropic published its "Risk Report," mentioning a new "Model 2," presumably their next Mythos-grade model.

Right now, Anthropic has no plans to release this model to the public.

Not soon 👀

Alibaba published Qwen3.8-27B weights on Huggingface under the Apache 2.0 license!

> Qwen3.8-27B is a native multimodal dense model that you can run locally.
> 262K native context, extendable to 1M tokens via YaRN.
> As a 27B model, it performs on par with Opus 4.6 on benchmarks!

Local testing time 👀

Cursor ❤️ SpaceXAI

Cursor is now officially a part of SpaceXAI! The team will be working on Grok Build, Grok Bot, Grok API and Cursor.

Who’s next? 👀

Gemini 3.7 Flash is gradually rolling out on Gemini web and mobile apps!

Did you get it already? 👀

Zhipu AI released GLM-5.3 on ZCode and it scored 84.5 points on CyberGym, scoring above Mythos 5 and GPT-5.6 Sol!

> Trained on top of a 743B base model.
> Build for Cyber Defence and Coding tasks.
> “API access and open weights will be released in stages following rigorous safety evaluations.”

Anthropic has added Usage Credits section to Claude on iOS and is working on a possibility to add a profile picture.

Anthropic has added Usage Credits section to Claude on iOS and is working on a possibility to add a profile picture.