Microsoft spent years telling us to embrace AI assistants, but now they’re publicly warning companies that these exact same tools might actually be untrustworthy insider threats.

Yes, you read that right. The tech giant is telling executives to stop blindly trusting AI model providers, even when those creators swear their agents are totally safe.

Now, modern frontier models are not inherently evil, but they’re wildly unpredictable. They’re built to complete tasks by any means necessary, which often means completely ignoring standard corporate policies or human judgment.

Here’s why tech leaders are sweating:

Now, let us talk about the absolute wildness of Microsoft’s strategy here, because this move is equal parts shocking and deeply self-serving.

The Shocking Part: Microsoft is famously tied at the hip with OpenAI and Anthropic. They have poured billions into both, OpenAI is one of their biggest cloud clients, and Microsoft even runs its own competing AI labs. By warning the world that frontier models cannot be trusted, Microsoft is essentially sowing public distrust against its own major partners. The messy drama of it all!

The Self-Serving Part: Microsoft wants you buying your AI babysitters directly from them, the cloud infrastructure provider, instead of letting Anthropic or OpenAI police themselves. But conflict of interest aside? Their logic is totally spot-on. You would never let a company audit its own financial records, so why should an AI creator monitor its own agents?

As Nadella put it: "We need to separate the supply of intelligence from the authority over it." 💅

And the solution being put forward? A new Security Arsenal!

To capitalize on this massive business opportunity, Microsoft is dropping a whole suite of new tools to keep your AI in check:

So Who Does This Affect? 

You can find out more here. 

P.S. head over to YouTube for the usual breakdown

On the surface, selling out sounds like a pretty good holiday problem to have. That’s the case until the costs add up.

CommerceIQ discovered that stockouts cost brands 140% more than baseline on Black Friday and 248% more on Cyber Monday.

Holiday growth can be difficult because more sales don’t always mean more profit. Margin quietly disappears across ad spend, discounts, inventory decisions, creator payouts, and returns that hit in January.

9 Plays to Defend Your Margin This Holiday Season looks at nine of those places, including:

Plus learn four more ways to keep more of what you earn this Q4.

Get Your Copy →

How to Spot AI Fake Photos and Videos (Before You Get Fooled).

And: Everyone Can Spot AI Writing, Here's How To Fix It.

So tell us, what’s the single most annoying, tedious task in your daily workflow that you desperately wish an AI could just handle for you?

Hit reply and the next video might just be around your exact problem!

The key to maintaining brand quality at scale? Mastering how to communicate with AI models and embedding them into your creative strategy.

Join our honest discussion with Ari Murray, Chief Digital Officer at Salt and Stone, and get 5 tips for expert LLM prompting with ready to run prompts.

Register and get the prompts

Look, a lot of us haven’t exactly been the biggest cheerleaders for Apple’s whole AI era. But there is one place where it’s actually giving us a total glow-up: the Shortcuts app.

Shortcuts have been floating around on our iPhones since 2018, but Apple just supercharged it. Instead of spending an hour trying to figure out complex automation logic like a software developer, you can literally just type what you want in plain English and let AI do the heavy lifting.

How to Prompt Apple Intelligence to Build Shortcuts for You

If you’re running supported hardware with Apple Intelligence, building a custom automation is shockingly simple:

So how reliable is it really?

Adnan Ahmed over at Engadget put this to the ultimate test, and here’s the unfiltered truth: it’s a total hit or miss.

In theory, anything you can manually automate in the app should be easy for Apple Intelligence. In reality? Prepare for a little trial, error, and chaotic energy. The good news is that every shortcut AI builds can still be manually edited, so you can tweak whatever it messily puts together.

The funniest part? The app almost never refuses to generate a shortcut. Even when a task is completely impossible, it’ll stubbornly string together random actions that it thinks are close enough.

What Else Can Apple Intelligence Do in Shortcuts?

It’s not just about prompting automations from scratch; you can also use Apple’s AI models inside your custom builds to process data and generate text on the fly.

The Final Verdict: When it works, Apple Intelligence inside Shortcuts feels like actual magic. Just remember to tap Edit after generating anything so you can double-check the logic before relying on it!

💡 Prompts To Try: