🚨 AI News | TestingCatalog

OpenAI announces GPT-5.5-Cyber (new) model update, which scores 85.6% on CyberGym benchmark in comparison to 81.9% in its early version.

Codex got a new Security plugin too 👀

ANTHROPIC 🔥: Claude for mobile is getting Cowork support soon!

Users will be able to trigger Cowork tasks on mobile and view scheduled tasks in the app.

> Keep Cowork going when you are on the go
> Start and steer tasks directly from your phone
> Check in from your phone, browser, or Claude desktop app
> Work continues in the background, even when you close the app

h/t DevMode
@testingcatalog

BREAKING 🔥: Sakana AI announced the Sakana Fugu and Sakana Fugu Ultra systems, which perform on par with Claude Fable 5 and Mythos 5 across many benchmarks.

> Sakana AI is an AI lab from Japan, and Fugu is an orchestration model trained to operate other LLMs.

> It is available as an API but not yet accessible in the EEA region.

That's a natural evolution. Orchestration multi-model systems will outperform single-model systems, and they will become much more accessible for smaller labs and companies to build.

Big players will have to consider building orchestrating systems that rely on models built by competitors. It is already happening at Meta, Apple, and Microsoft, and will likely catch Google, Anthropic, and OpenAI as well eventually.

ICYMI 👀: Cursor got a new /automate Skill

Automation your toil got insanely simpler over the past few years with AI.

Even Automation is Automated now 🤖

Grok Build Remote appears to be accessible on the web. However, it is not functional and it is likely unintended.

Both Grok Build web and desktop apps are now under a big question, if they would survive Cursor acquisition or not.

Link below 👀

Meta AI is getting a new Artifacts tab on the web. All the presentations, docs, web pages and other creations would be stored over there.

Bridging the feature gap 👀

ICYMI 👀: Users can now search for Imagine images and videos on Grok!

It works like a proper image search but scoped to your Imagine creations. I wish we have it across the whole set of published Grok images eventually - that would be huge!

Anthropic is working on "Schedules" for its upcoming Claude Conway.

> Recurrent triggers that wake Conway on a schedule. Survive container restarts.

Super excited to see how Conway will work with all these planned features.

Google is working on a new Artifact type for NotebookLM called "Lit review". In this mode, NotebookLM will be able to "Generate a Literature Review Matrix" based on your sources.

Considering upcoming additions of Google Play Books and Text Books as sources, Google is planning to push new use cases for readers and writers.

Will it be able to map out all characters from "A Song of Ice and Fire"?

Perplexity releases Brain Memory System for Perplexity Computer

ICYMI: Perplexity’s Brain is a shared memory layer for Search, Computer, and possibly Comet, exposing organized topics, source context, and a 3D knowledge map. It reportedly lifts correctness, recall, and task cost where prior context matters.

🗞 #perplexity @testingcatalog

Anthropic launches live Artifacts for Claude Code

ICYMI: Anthropic launched Artifacts in Claude Code, letting Team and Enterprise users turn coding sessions into live, shareable private web pages. It supports real-time updates, version history, and org-level access controls for technical teams.

🗞 #claude @testingcatalog

Anthropic launches managed connector access with Okta

ICYMI: Anthropic added enterprise-managed authorization for MCP connectors, starting with Okta. Admins can provision and revoke connector access through their IdP, granting teams first-login access across Claude tools with support for major providers.

🗞 #claude @testingcatalog

OpenAI is preparing to add Library from ChatGPT into Codex as well.

The next phase of consolidation 👀

Mistral AI released Code on Vibe to Pro users. Desktop app has been confirmed as well.

Soon 👀

ICYMI 👀: Z AI aims to reach Fable 5 level already this year.

xAI wen?

Claude Enterprise admins can now centrally authorize MCP connectors for their organizations via a new Enterprise-Managed Auth extension.

Mass MCP 👀