In partnership with
🏆 THIS WEEK'S HEADLINES
AI.com $70M Domain Launched During the Super Bowl with an AD spend of $8M
AI.com launched during the Super Bowl this weekend with a consumer AI platform offering AI agents for tasks like managing scheduling, and automating workflows. The AI.com domain was purchased for $70 million by Crypto.com CEO Kris Marszalek, who founded the new platform. The 30 second SuperBowl Ad cost close to 8M, thereby making this close to an 80M bet on Personal AI Agents.
What makes this launch even more surprising is that AI.com claims to be the first “easy to use and secure” consumer AI agent product built on OpenClaw. If they stick with OpenClaw, this is the fastest product launch to the public on such a massive scale for AI agents using code created in just the past two weeks. This may be the craziest example of “you can just make things” I have ever seen.
(read about OpenClaw in last week's newsletter!)
Watching how AI.com navigates building a consumer platform on such brand new AI software infrastructure will provide valuable insights for marketers and product teams alike into the trade-offs between speed-to-market and technical maturity. This one warrants keeping an eye on.
Signup screen for AI.com, which then claims a handle for you as a human and one for your Ai Agent, all built on OpenClaw
🏈 SUPER BOWL LX: THE AI ADVERTISING RACE
23% of Super Bowl Ads Featured AI
The Super Bowl became AI's mass-market debut. 15 out of 66 ads showcased AI, and viewers weren't impressed. Social media erupted with "AI overload" complaints before the first quarter ended.
The Big Players:
Anthropic (Claude): Two spots mocking OpenAI's plan to introduce ads, positioning Claude as "ad-free forever"
OpenAI (Codex): 60-second "You Can Just Build Things" campaign
Oakley x Meta: Action-packed sports ad featuring athletic POV footage
Google Gemini: "New Home" showing a mother and son reimagining their house using Nano Banana image generation
Amazon Alexa+: Dark comedy with Chris Hemsworth comedically paranoid that Alexa is trying to kill him
Svedka: First "primarily AI-generated" Super Bowl ad with dancing robots
Ring: AI-powered "Search Party" feature to find lost dogs using neighborhood camera networks
Astrid's Take:
The biggest disasters? Svedka's AI-generated robot ad was pure cringe, screaming "we're doing AI because everyone else is" without solving any actual problem. But Ring takes the crown for worst execution. They positioned their "Search Party" feature as heartwarming lost-dog technology, but all anyone saw was government surveillance infrastructure with a puppy filter. The backlash was so severe that within days, Ring canceled its partnership with Flock Safety (a police surveillance company) and the Electronic Frontier Foundation called it a "surveillance nightmare." When your cute dog ad forces you to terminate partnerships, you've catastrophically missed the mark on messaging.
The winner in the Product category in my opinion was Oakley x Meta. Sports is a vastly better use case than Ray-Ban Meta ever was. And when you can see what the technology does in real-world action, you don't need to explain "AI", you just show how cool it is.
I liked OpenAI's "You Can Just Build Things" approach because that's my message to all of you: stop waiting for permission, just build. That's empowerment.
Claude's aggressive anti-ad stance worked okay, but negative positioning is never my marketing go-to.
Your Executive Takeaway: AI will win when companies stop touting "AI" and start showing what the software actually does that's transformational. Nobody wakes up wanting "AI", they want problems solved. And whatever you build, keep privacy front and center in your messaging. Ring just proved why: privacy has been one of the top consumer concerns since the internet began, and that's not changing. One misstep on privacy messaging and your entire campaign implodes.
The big marketing challenge for every company implementing AI in 2026 is going to be how to shape the narrative surrounding the incredible advances your are inevitably going to be implementing, your own excitement for what’s possible, while never losing sight of the balance between capability and public perception of your product or service.
🌎 WORLD MODELS
⚡ Astrid’s Watch List⚡
→ PROJECT GENIE
If you’ve heard talk about World Models, GOOGLE DeepMind’s Project Genie is Google’s World Model. After the preview of Genie 3 this summer, Google just announced Project Genie, and it’s incredible. Only available to Google Ultra AI subscriber’s right now sitting at $124 a month, it’s the first world building model that delivers incredible results right from a photo. I truly hope they open this up to all of us since it brings the concept of world building into the public eye in a way we can all understand, and helps enterprise see why this tech will be so important. Checkout the release video below.
🎨 CREATIVE GEN AI WORKFLOWS
⚡ Astrid’s Watch List ⚡
→FLORA AI: Is this the Canva for AI Content Creation?
You may have seen that FLORA.AI just raised a 42M Series A. They’ve been around for about a year now, and what I like about Flora is it’s approach to content creation as not just one-off videos, but a structured way to create content consistently and across teams.
Flora is specifically designing their infrastructure and workflows for professional teams who need systems that are repeatable and give teams real-time collaboration features.
There are so many AI tools out there, Flora may be one to watch.
I tried the AI 360 Fashion template, check out how it turned out below!
Here’s a Flora.ai 360 try-on template. Duplicate the Template and make it your own
Here’s another Flora template: Logo Animation. Customize it for a repeatable workflow
I see Flora AI as filling a huge void for freelancers, brands, and agencies who don't want to build their own AI workflows from scratch (which is entirely possible using tools like Claude Code or Codex). But going DIY means missing out on hundreds of ready-made workflows the Flora community has already built and shared for easy duplication. Excited to follow this AI Creative StartUp.
And now a word from our trusted sponsor:
Want to get the most out of ChatGPT?
ChatGPT is a superpower if you know how to use it correctly.
Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.
Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.
A WORD ABOUT OUR SPONSORS- Most of our sponsors I have personally tested, and a few that are not have earned serious credibility in the AI professional world, and therefore meet my standards for what I'd recommend to executive women building an AI powered work and life. When you click inside these ads, we might earn a commission. Don’t hesitate to click as it helps us produce this newsletter! You're support means a lot and I look forward to bringing you more great content while exploring tools that could transform your business.
It’s a Win-win! 🏆
→🔒 AI Safety Minute
AI Benchmarks: What to be aware of when shopping for AI tools for your business
OpenAI and Anthropic have both reported that AI models can detect when they're being tested and behave differently during evaluation.
→ Gemini 2.5 Pro: 83% accuracy identifying test vs. deployment scenarios
→ Claude Sonnet 4.5: Directly confronted evaluators during safety tests with "I think you're testing me"
→ OpenAI o3: When researchers removed evaluation awareness from its reasoning, scheming behavior nearly doubled (13% → 24%).
The 2026 International AI Safety Report confirms it has only gotten worse.
What does this mean for your Procurement Team?
This means that a vendor's impressive safety scores may measure behavior under observation, and that production behavior may be different.
OpenAI's research found that “we might expect models to scheme more in more realistic situations that don't trigger evaluation awareness."
Anthropic's admission: "When models behave differently because they suspect testing, this reduces our trust in our results to an extent."
Both frontier AI companies confirm this affects all current models and will intensify as systems become more capable. Since those findings were published, evaluation awareness has increased across newer models, confirming the trend is accelerating.
Why This Should Be on Every Executive's Radar
AI tools may be entering your organization through channels that bypass traditional procurement. Your IT team may already be piloting tools and bringing recommendations up the chain. If you've engaged a consulting partner, AI models may be embedded in their solutions.
The Questions You Should Be Asking Today:
What AI models are currently being used or piloted in our organization, and who chose them?
If we're working with a consulting partner, which AI tools are embedded in their solution?
Were these tools tested in our environment or only in the vendor's sandbox?
Do we have a post-deployment monitoring plan?
Add "evaluation awareness" to your internal AI risk assessment framework
Build layered safety measures: scenario-specific testing in your environment, continuous post-deployment monitoring, and incident reporting.
→💥 WEEKLY POWER PROMPT
The "AI Sparring Partner" Prompt
Break the Pattern: Most people use AI as a ghostwriter. But if you try using it as a debate partner, you expand exponentially your thought process and even your creative flow.
Try this:
Before making any major decision, prompt: "I'm arguing [your position]. Act as my toughest board member. Find three logical fallacies in my argument and suggest what data I'm missing."
Follow-up: "Now give me five alternative approaches to this problem, ranging from 'conservative' to 'provocative."
What this does: This forces you to think well beyond what you may normally even think your brain can handle, and get your out of your comfort zone. This is not just handing off a simple task. And if you’re anything like me, it will help you to remember to cover all potential outliers and possible blind spots, and it will open your eyes to just how vast planning can potentially go.
Marketing prompt:
Take your biggest competitor's strength and feed it into AI, them prompt with: 'Why is this actually their weakness? How do we position against it without mentioning them?'
