Need to Know News - February 26th, 2026
In this week's Need to Know News edition:
🤖 A sharper, faster Nano Banana 2 launched today... and its keeps up to 5 subjects consistent across multiple images for seamless ad campaigns.
🤖 Perplexity launched a digital worker that runs entire workflows for hours or months... this isn't a chatbot anymore.
🤖 Online fashion returns could hit $850 billion this year... one AI agent trained on 20 years of purchase data...across 91,000 brands...is attacking the root cause for your brand.
And a whole lot more!
Nano Banana 2 Is Out: Combining Pro Capabilities With Lightning-Fast Speed
While Google's original Nano Banana model went viral last August, its Pro version stayed locked behind premium tiers. Nano Banana 2 changes that equation. It packs Pro-level smarts, subject consistency across five characters, and resolutions up to 4K into Flash speed.
The model rolls out today across the Gemini app, Search, Google Ads, and Flow. Text rendering and translation work inside images now too.
Perplexity 'Computer' Orchestrates Every Top AI Model Into One System
Forget building around a single AI model. Perplexity's new system, Computer, treats Opus 4.6, Gemini, ChatGPT 5.2, Grok, and others like specialized employees on a shared team. Describe an outcome, and Computer breaks it into tasks, spawns sub-agents, and runs them in parallel for hours or even months.
Each sub-agent gets a real browser, a real filesystem, and real tool integrations. If one hits a wall, it creates more agents to solve the problem before checking in with you. Available now for Max subscribers.
🚀 Every Ad, Landing Page, & Offer Need these 5 Crucial 'Copy Blocks'
Two $100+ million copywriters discovered what REALLY drives conversions...5 simple building blocks arranged in a specific order. This framework is so powerful that beginners are generating 5-figure profits with it. And you can learn the full system in just 33 minutes.
Discover All 5 'Copy Blocks' Now
True Fit's AI Agent Tackles Fashion's $850B Returns Problem
Seventy percent of questions shoppers ask AI agents about fashion boil down to one thing: "Will this fit?" True Fit built an agent to answer that, drawing on 20 years of purchase data from hundreds of millions of shoppers across 91,000 brands.
It watches for hesitation on product pages, then delivers size guidance based on what people actually bought and kept. No static charts.With projected returns hitting $850 billion, retailers leaning on generic data are bleeding money at the source.
Meta Courts Independent Agencies to Fund Its AI Ambitions
Meta needs more ad revenue to bankroll its AI ambitions. So it's courting independent agencies again, for the first time in years. The new program, Agency Growth Collective, swaps the old single-rep model for a rotating bench of Meta specialists and virtual training sessions.
It's U.S. only for now, with no minimum spend requirements announced, though participation gets monitored. Agencies that don't engage risk getting cut. While it mind sound generous, the reality is a pipeline to bigger brands and bigger budgets.
AI Is Upending Marketing on Two Fronts
Online searches drop roughly 20% after people adopt ChatGPT, according to London Business School and UCLA researchers. The highest-value customers shift fastest. Brands built on information delivery face the steepest risk, while those rooted in community have natural armor.
But a second shift cuts deeper. AI agents are starting to make purchase decisions alone, and different models behave differently. GPT favors the first product slot on a page. Claude picks the middle. Gemini leans right. Marketers now need bot psychology alongside human psychology.
Burger King's AI Headset Listens for "Please" and "Thank You"
Five hundred Burger King locations are strapping OpenAI-powered headsets onto employees, and the AI does more than answer recipe questions. The system, called Patty, flags low inventory, removes sold-out items from digital menus, and alerts managers to dirty restrooms in real time. That's the operational pitch.
The stranger part: Patty also tracks how often workers say "welcome," "please," and "thank you," feeding those patterns to managers as hospitality signals. Burger King insists it's coaching, not surveillance.
Full Story
Smartminded Analyzed 600+ Amazon Listings for AI Content… The Results Were Surprising
Detecting AI-written copy on Amazon turns out to be messier than expected. Researchers ran 500 listings through two detection models and found roughly one in five showed strong AI patterns. Another 40% landed in a gray zone.
The sweet spot? Products priced $20 to $50, where competition runs hottest. Longer listings scored higher for AI signals but correlated with fewer reviews and lower ratings.
If you're an Amazon seller, the takeaway is that the algorithm doesn't care. Rankings showed zero meaningful difference between AI and human copy.
68% of TV Producers Prefer AI-Optimized Story Pitches
Here's a number that should stop every PR team mid-pitch. 68% of local TV news producers say they're more likely to air a story when it's been optimized for AI search. Not slightly more interested. Significantly more.
Tnew report from D S Simon Media surveyed producers nationwide and found 37% already use AI tools to find stories worth covering. 60% of stations now optimize their own content for AI discovery too.
Airbnb's AI Now Handles a Third of Customer Problems Without a Human
Airbnb quietly slipped an AI assistant into customer support across the U.S., Canada, and Mexico. It already resolves about a third of all issues before a human agent ever gets involved. Resolution times dropped sharply. The company plans to roll it out globally later this year.
But that's just the warm-up. Airbnb is also testing AI-powered search that lets travelers describe trips in natural language instead of clicking through filters. Brian Chesky called it "scaling AI with discipline." Translation: they're moving fast, but not saying so out loud.
Full Story
Google Adds Automated Workflows to Opal
Google just handed its vibe-coding app Opal a new trick: automated workflows. Users describe a task in plain language, and a Gemini 3 Flash-powered agent figures out which tools to use, maintains memory through Google Sheets, and plans each step on its own. No coding required. If the agent needs more info, it asks.
Opal first launched for U.S. users in July 2025 and now operates in over 160 countries, competing directly with Lovable, Replit, and a growing pile of natural-language app builders.
Thanks for reading.
Until next time!
The AI Marketers
P.S. Help shape the future of this newsletter – take a short 2-minute survey so we can deliver even better AI marketing insights, prompts, and tools.
[Take Survey Here]
