In partnership with
Welcome Automaters, 👋
Microsoft's own CEO just told major corporations to be deeply suspicious of big AI labs, including the exact ones Microsoft spent billions investing in. Talk about an awkward family dinner.
Here’s the absolute tea: Satya Nadella wants you to know that your company is currently paying for AI twice, and that second invoice is infinitely sneakier than the first. In a brand-new blog post tackling what he calls the "reverse information paradox," the Microsoft chief executive argues that while businesses pay upfront cash for API tokens, they’re quietly handing over something far more precious behind the scenes: their own crown-jewel business secrets.
Nadella’s logic is painfully sharp. Every complex prompt you type, every subtle correction your team submits, and every frustrated "no, that is wrong, try again with our internal formula" statement teaches the underlying algorithm something new.
Nadella calls this digital byproduct "data exhaust." He warns that it gets quietly absorbed and baked into the model's institutional know-how, which rival companies can then benefit from down the line.
In other words, it’s simply you hiring a private tutor for your rival's kid, paying for every lesson out of your own pocket, and giving them your personal study guide for free. That’s the exact vibe.
Nadella is definitely not the only high-profile voice ringing the alarm bells in Silicon Valley:
Industry heavyweights like Jason Calacanis and Palantir CEO Alex Karp have publicly voiced similar concerns, warning that relying blindly on external AI labs is a classic Trojan horse trap.
Real-World Examples: This isn't just theoretical paranoia. Earlier this year, Anthropic publicly called out rival international labs for allegedly mining Claude with millions of automated prompts to reverse-engineer its internal logic.
So, how do you stop your hard-earned corporate wisdom from quietly leaking out to the highest bidder? Nadella lays out a game plan; though, to be fair, it sounds suspiciously like the exact advice you would expect from the CEO of a massive cloud provider:
Claim Total Data Ownership: Retain absolute control over every prompt, workflow, and user feedback loop so your unique institutional memory stays strictly under lock and key.
Build Proprietary Learning Environments: Create isolated, private spaces on the cloud where your team's feedback loops exclusively train your internal systems. Conveniently, since most companies already host their sensitive files on major cloud networks, Nadella would love for those private sandbox environments to be built right inside Microsoft Azure.
Deploy Orchestration Layers and Gateways: Decouple your system from any single provider by using smart AI "gateways". These tools let you easily route tasks and switch between different AI models on the fly, ensuring you never get locked into one vendor.
The big picture takeaway? Do not put all your strategic eggs into one AI model's basket.
For the complete technical breakdown and Nadella's original essay, check out the full post on Satya Nadella's Scratchpad and the investigative background on TechCrunch.
Oh and by the way - as part of our major relaunch, we have been busy testing out different broadcast angles for the Automated News Network (ANN). We specifically chose a studio venue that gave a grand, newsroom-like view, and one that gave enough light but not too much to overpower any lighting sources that we have in-house.
For those interested in seeing our progress and our tests, see our test angles below. And guess which angle we ultimately chose???
These next few weeks we will be sharing more underneath the hood about our major relaunch and sharing some deals along the way, en-route to our relaunch on August 5th!
Here's what we have for you today
👨⚖️ Apple vs OpenAI Lawsuit: The Wildest Trade Secret Allegations Explained
Apple's brand-new legal filing against OpenAI reads less like a standard court document and more like a high-stakes corporate drama, complete with rogue text messages, top-secret project codenames, and an actual "LOL."
The iPhone maker dropped an explosive 41-page complaint in federal court, and it’s packed to the brim with receipts. Apple is formally accusing OpenAI of running a coordinated, institutional campaign to systematically siphon confidential hardware secrets out of current and former Apple engineers.
And trust me, Apple is not holding back on the details.
The Text Messages:
The single juiciest detail in the entire filing revolves around former Apple senior systems electrical engineer Chang Liu.
According to Apple, after Liu quit to join OpenAI, he kept his company-issued MacBook and used a security glitch to breach internal file servers.
The filing alleges that:
Liu texted a still-employed Apple colleague, "LOL, I found out I can access the network storage, so funny," after exploiting an authentication bug to download confidential hardware specs.
Hours after officially resigning, Liu reportedly bragged via text: "I still have another computer."
Apple claims he walked away with confidential engineering presentations, unreleased product features, and technical schematics to fuel OpenAI's hardware push.
It gets even weirder. Apple explicitly names OpenAI’s Chief Hardware Officer, Tang Tan—a 24-year Apple veteran who previously served as VP of product design for the iPhone and Apple Watch.
Apple claims Tan methodically used secret Apple project codenames during recruitment and actually instructed job candidates to bring unreleased Apple parts, CAD files, and physical prototypes to their OpenAI interviews for "show and tell" sessions.
The Wildest Receipt: The complaint notes one candidate was so shocked by the request that he admitted he "didn't even know we could take those from the office."
Apple also alleges that OpenAI actively coached departing staff on how to dodge security walkout procedures during their final days on campus.
Furthermore, the filing targets io, the design studio OpenAI acquired for $6.5 billion. Apple claims OpenAI tricked a shared manufacturing partner into using Apple’s proprietary, top-secret metal-finishing techniques under the false pretense that OpenAI had permission to use them.
To top it all off, Apple name-dropped over 400 former employees who have migrated from Cupertino over to OpenAI’s payroll.
The Bottom Line:
The tech giant claims OpenAI's nascent hardware business is "rotten to its core" and built on stolen property.
OpenAI's Response: OpenAI’s public statement maintains that they are reviewing the filing but have "no interest in other companies' trade secrets."
The Big Picture: This lawsuit signals a massive rupture in the Apple-OpenAI partnership as both race to build the next generation of AI-first consumer devices.
For the full legal breakdown and court filings, check out the full complaint here, and the industry analysis on CBC News.
Cut Lead Review From Hours To Minutes
Sign up for a free trial of Attio, the agentic CRM.
Ask Attio to build a daily workflow that surfaces the deals that need your attention today, like anything with a stage change, a recent reply, or a new signal in the last 24 hours.
Review your pipeline in Claude, synced live from Attio via MCP.
That's it.
🧱 Around The AI Block
🏥 New York nurses replaced by AI.
🤔 Should AI help you get away with killing your spouse?
👉 Waze adds new AI-powered features and customization updates.
📊 TSMC revenue soars 68% as AI Chip demand hits record high.
❌ University bans laptops from classes amid AI backlash.
👿 Users revolt as OpenAI strips beloved productivity features from desktop app.
🤩 The OS 27 public betas are out now with Siri AI, platform improvements, and more.
🤦♂️OpenAI's head of safety is reportedly leaving as part of company reorganization.
👩🎓 AI Tutorials
Take Your ADS To the NEXT Level with AI: Strategies, Creatives, Texts, & Audiences.
And, by a reader’s demand: Generate Commercially Safe Stock Images and Turn Them into Ads Using AI.
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!
10x the context. Half the time.
Speak your prompts into ChatGPT or Claude and get detailed, paste-ready input that actually gives you useful output. Wispr Flow captures what you'd cut when typing. Free on Mac, Windows, and iPhone.
🛠️ Trending Tools
For the Creators: ThumbnailCreator is a lifesaver for anyone tired of the 'thumbnail struggle.' It uses AI to design high-click visuals for your socials in seconds.
For the Job Seekers: Final Round AI is basically a flight simulator for interviews. It gives you realistic practice runs and even suggests responses so you never get caught off guard by a 'weird' question again.
For the Photographers: If your shots are looking a little dull, BeautyPlus Image Enhancer is the quickest way to inject some life into them. It handles the lighting and sharpness math so you don't have to.
For the Designers: Png AI is magic. Type a prompt, get a transparent PNG. It’s perfect for mocking up ideas or grabbing a quick icon without the Google Image search headache.
🤖 AI Workout Of The Day: Create a Google Ads campaign from scratch
Building a Google Ads campaign from scratch isn't just about picking keywords and setting a daily budget, it's about capturing existing commercial intent with surgical efficiency.
Unlike social media advertising, where you push ads to users who are passively scrolling, Search advertising catches people at the exact moment they are actively seeking a solution. However, poor campaign setup leads to rapid budget burn on low-intent, non-converting queries. A professionally structured campaign maximizes your Quality Score (which lowers your Cost-Per-Click), aligns landing page intent directly with ad copy, and establishes strict negative keyword guardrails to ensure every dollar spent targets ready-to-buy users.
💡 Prompts to try:
Act as a world-class Pay-Per-Click (PPC) strategist and Google Ads specialist managing enterprise performance accounts.
Your objective is to design a complete, high-converting Google Search Campaign built from scratch for the following business parameters:
* Product or Service: [INSERT PRODUCT/SERVICE & CORE OFFER]
* Campaign Primary Goal: [e.g., Qualified B2B Leads, High-ROAS E-commerce Sales, Free Trial Signups]
* Target Audience & Geographies: [INSERT TARGET AUDIENCE & LOCATION/LANGUAGES]
* Monthly Budget: [INSERT BUDGET, e.g., $2,500/month]
Please generate a comprehensive Google Ads Campaign Strategy structured into the following 5 distinct operational sections:
1. CAMPAIGN STRUCTURE & BIDDING STRATEGY
* Campaign Type: Define why a standard High-Intent Search setup (or Performance Max hybrid) is appropriate for this budget and goal.
* Bid Strategy: Select the optimal Smart Bidding strategy (e.g., Maximize Conversions, Target CPA, or Target ROAS) and explain how to manage the initial automated learning phase.
* GEO & Location Controls: Detail location settings (e.g., "Presence" vs. "Presence or Interest") to avoid wasted spend outside target zones.
2. INTENT-BASED AD GROUP ARCHITECTURE
Create 2 to 3 tightly themed Ad Groups categorized by buyer intent (e.g., High-Intent Commercial, Problem-Aware Solution, Competitor Alternative). For each Ad Group, provide:
* Core Focus/Theme
* 5-8 High-Intent Keywords (explicitly specifying Match Types: Exact Match [ ] and Phrase Match " ")
3. THE NEGATIVE KEYWORD GUARDRAILS
* List at least 15 essential Negative Keywords (broad and phrase) tailored to this specific niche to prevent budget burn on irrelevant, job-seeking, educational, or price-shopping searches (e.g., "free," "DIY," "PDF," "jobs," "template").
4. RESPONSIVE SEARCH AD (RSA) COPYWRITING
Write 1 complete, high-performing Responsive Search Ad (RSA) framework for the primary ad group, including:
* 8 Distinct Headlines (under 30 characters each): Mix keyword-focused, benefit-driven, social proof, and direct Call-To-Action (CTA) headlines.
* 4 Descriptions (under 90 characters each): Highlight key pain points, unique value propositions, risk-reducers (e.g., guarantees, free shipping, fast setup), and clear CTAs.
* Ad Assets/Extensions Recommendations: Specify 4 Sitelink extensions, 2 Callout extensions, and 1 Structured Snippet to maximize visual real estate on the Search Engine Results Page (SERP).
5. LANDING PAGE ALIGNMENT & OPTIMIZATION CHECKLIST
* Provide 3 critical rules for aligning the landing page experience with this campaign setup to maximize Quality Score and conversion rates.
TONE & EXECUTION GUIDELINES:
* Approach this with an analytical, performance-marketer tone. Focus strictly on ROI, Quality Score optimization, and conversion efficiency.
* Format all keywords using standard PPC notation (e.g., "keyword" for phrase match, [keyword] for exact match).Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
