AI moves fast.
Know what matters.

What’s changed, why it matters to your work, and something useful to try next.

Latest stories

224 stories for work · Last 180 days
More filters
AI provider

By default, we show stories relevant to leaders, managers and business owners.

MistralModel releases

Mistral OCR 4: SOTA OCR for Document Intelligence

Relevant to anyone who deals with scanned documents, forms or PDFs, including in languages other than English. It pulls out structured text with a confidence score for each section, which is useful for digitising paperwork or records more reliably than older OCR tools.

PerplexityNew features

Deep Research, command panel, and enterprise controls

Useful for anyone who needs to turn a broad research question into a finished output — the tool can now take a multi-step research task and produce a report, spreadsheet, dashboard or simple website directly, saving the step of manually reformatting findings. Admins also get finer control over how AI credits are spent per team member.

xAINew features

Grok for Word

If you write reports, proposals or briefing documents, this free Word add-in turns a prompt into a structured draft, pulling in web/X research plus content from your connected email and cloud storage. It effectively drafts a first version of a document from a brief description, rather than you starting from a blank page and researching separately.

OpenAINew features

Scheduled tasks in ChatGPT

If you set ChatGPT to run a task on a recurring basis (like a daily summary), there's now a dedicated Scheduled page so you can see and manage all of them in one place, and notifications about completed runs are more reliable, so you'll actually know when something has run.

GoogleNew features

June Pixel Drop: New features for creators, Gemini upgrades and more

Relevant to Pixel phone owners on a paid Google AI subscription who want to create or edit short videos and generate music through conversation rather than dedicated editing software. It is a content-creation feature for personal or social use, not a general work-productivity update, and is restricted to adults.

MicrosoftNew features

Anthropic's Claude now available in Microsoft 365 Copilot

For anyone using Copilot Chat who wants a second model to choose from. You can now select Anthropic's Claude instead of the default when working on complex analysis, digesting a long document, or drafting structured content — useful if you want an alternative take on the same task.

xAIModel releases

Grok Imagine Video 1.5

Relevant if you already use Grok Imagine to generate video content — this update improves the realism of motion, physics and audio, and adds Projects plus multiple agents for managing several video jobs at once. Mainly useful for people producing marketing or social video rather than general office work.

xAINew features

Grok for PowerPoint

If you regularly build presentations, this free PowerPoint add-in turns a rough outline into a full slide deck complete with web/X research, diagrams and images already inserted. Instead of starting from a blank slide and hunting for supporting content yourself, you describe the deck you want and get a first draft to refine.

AnthropicPolicy

Statement on the US government directive to suspend access to Fable 5 and Mythos 5

Relevant mainly if you or your organisation were actively relying on Fable 5 or Mythos 5 for work — this explains why access to both suddenly disappeared worldwide (a government export-control order over a possible jailbreak risk), so any workflows built on those specific models would have broken without warning. If you were not using those two models, there is no practical impact.

OpenAIRetirements

Retiring GPT-5.2 models

Mostly invisible housekeeping — if you had conversations running on the older GPT-5.2 models, they've automatically carried on using the newer GPT-5.5 model instead, with nothing for you to do.

AnthropicModel releases

Claude Fable 5 and Claude Mythos 5

Fable 5 is the one most people will actually get to use — a new top-tier general model with strong safety guardrails, useful for the same everyday work as any Claude release (writing, analysis, coding). Mythos 5 is not for general use: it is deliberately limited to vetted cybersecurity partners and has fewer safeguards, so it is only relevant if you work in that specialised security field.

Steve Wiggers, SCALA AI trainer and speaker

Useful context. A clear source.

SCALA’s updates draw on published provider announcements. Each article separates the source summary from practical interpretation. Steve’s guides and the SCALA method help you take the next step.

Meet Steve and SCALA