SuperMarketers

If you are a High Growth B2B SaaS, SuperMarketers Is Your Done-For-You Demand Gen That You've Been Needing. 

Our primary goal is to boost your SaaS revenue effortlessly with our expert, done-for-you demand generation services—saving you stress and costs compared to running an in-house marketing team.

Book Your Free 45 Minute Strategy Session To Learn How To Use AI To Scale Your SaaS More Profitably and Efficiently: 

👉 https://tidycal.com/genfurukawa1/45-min-marketing-strategy-session-with-gen-furukawa

Founder

MarketerHire

Course Creator and Lead for MarketerHire's first AI course. 

The course is a 30 Day Intensive on Content & SEO, teaching marketing experts and freelancers in the MarketerHire community.

Marketer in Residence

Content at Scale

Joined Content at Scale as VP of Marketing to build brand awareness in the AI writing space. 

In my short time there: 
- Built Content at Scale's Prompt Library -- created 10k prompts to help entrepreneurs, marketers, business leaders learn how to use AI to accomplish their goals. 
- Helped launch a community for Content at Scale. Educated members on AI use cases for marketing with weekly workshops, webinars, and tutorials. 
- Recorded 14 podcast episodes on AI and Marketing with industry leaders like Brian Dean (Backlinko, Mike King (iPullRank), Travis Jamison (Investing.io).

VP Marketing

Prehook

I co-founded and sold Prehook, a quiz platform for Shopify merchants. 

Our goal was to build a software to help Shopify merchants learn more about their customers with quizzes. 

In just a few minutes, merchants could publish a quiz to ask a few questions, capture an email/SMS, and recommend a product. No coding required. 

Brands like Spikeball, Smile Direct Club, Evolve Skateboards, JOI, Teelixir, Camille Beckman and hundreds grew their audience, captured zero party data, and drove incremental revenue with Prehook's quizzes. 

Created thought leadership around ecommerce quizzes and zero party data in:
- our blog
- weekly newsletter
- podcast (CartOverflow.com)
- industry events (2x eCommerce, AdWorld, ProductLed Growth Summit)
- Guest on industry podcasts (eCommerce Fastlane, $100 MBA, eCommerce Badassery etc)

Co-Founder

CartOverflow.com (podcast)

Learning and sharing ecommerce marketing with smart people doing cool things. Weekly episodes with interesting guests (operators, agencies, service providers, SaaS folks) alongside co-host Jeremy Biron. episodes at www.cartoverflow.com

Co Host

Retainable

Retainable helps B2B companies reduce churn.

Boost your monthly recurring revenue, instantly. Retainable helps B2B software companies create a flywheel — increased retention; higher lifetime value; higher MRR. Email me if you have questions: gen@retainable.com

Jungle Scout

On the founding team and first marketing hire, helped lead growth of the company to over 100k customers. 

Working alongside CEO and executive team, I am responsible for the marketing strategy, recruiting/hiring, and execution of all Jungle Scout products. This includes Jungle Scout, Jump Send, Splitly, and Fetcher. These marketing efforts have helped establish Jungle Scout as the market leader for all Amazon seller software and educational resources. 

Particular focus on customer acquisition and retention. The primary marketing responsibilities include growth from SEO, content, paid, social, email, copywriting, conversion rate optimization, YouTube and more.

Helped lead Jungle Scout's international marketing strategy, particularly the growth in the Chinese market. This includes recruiting and hiring the first marketer focused solely on the Chinese market, and establishing the Go To Market strategy, junglescout.cn site, and helping to manage the growth of the team in China. These efforts have contributed significantly to Jungle Scout's Annual Recurring Revenue.

Vice President of Marketing

Amazon Product Research, Made Easy. 

Leading the digital marketing efforts for the industry-leader in Amazon product research, I focus on strategy and execution of: content, email, SEO, paid acquisition, analytics, split testing, conversion rate optimization, copywriting, and more.

Head Of Marketing

Fetcher.com

Growth

Splitly

Splitly is the only tool to help Amazon sellers maximize profits by running automated split tests.

Jump Send

Building a marketplace for the top Amazon Email Automation and Coupon Distribution Tool. I focus on helping sellers leverage email follow up campaigns to gather product reviews and improve conversion rates.

Done-For-You B2B SaaS Marketing | Founder @ SuperMarketers: Generate More Leads Effortlessly | Let Proven Experts Handle It All | Book A Call To Learn How 👇

Are you a SaaS leader looking to improve your marketing -- more high-quality, inbound leads at scale?  

SuperMarketers is a Done-For-You marketing agency that handles everything: Strategy, Execution, Marketing Ops, and Growth. 

We can do this more efficiently, and with greater impact, with our proprietary AI-driven marketing system. 

At a fraction of the cost of growing your in-house team or working with traditional agencies. 

I can do this because I've been in your shoes. 

I was a bootstrapped co-founder of an eCommerce SaaS, called Prehook. 

We sold in April 2023. 

It was a six figure sale. 

I know that it would have been a bigger outcome if I had the SuperMarketers growth system. 

I am deep in the AI and marketing space: 

I teach AI and marketing (for MarketerHire, a global network of top tier digital marketers). 

I podcast about AI and marketing (podcast is called SuperMarketers of course).

Helping write a book on AI and marketing (with Jon Nordmark, CEO of Iterate.ai).

And of course I help B2B SaaS brands grow profitably leveraging AI and automation--market leaders like Octane AI, Iterate.ai, Pipio.ai, and more.  

Previously I was part of the founding team and VP of Marketing at Jungle Scout, leading a team of 25 marketers through the $110M fund raise lead by Summit Partners. 

Book a 1 on 1 Strategy Call With Me To Learn Exactly How You Can Use Ai To Scale Your Marketing More Profitably and Efficiently, Without Spending Any Money. 

➡️ Book It Now: https://tidycal.com/genfurukawa1/45-min-marketing-strategy-session-with-gen-furukawa

🌷 Gen Furukawa

Ross Hudgens analyzed 200 websites to see what's working in search.

1️⃣ ChatGPT traffic is flattening. 

2️⃣ Google traffic is growing.

But the type of content that ranks has completely changed.

For most B2B companies, pipeline is steady or growing even if topline traffic is soft.

The shift is in what kind of content actually drives results, specifically: 

✅ Guides and playbooks still get the most traffic.
✅ Comparison pages are holding strong.
✅ Pricing and cost pages are performing well.
✅ Case studies saw a lift in engagement.
✅ Trend content is getting attention.

What's dying:
🟥 "How to" posts are losing visibility.
🟥 "What is" definitions are getting buried.
🟥 Anything that can be answered in a single snippet isn't getting clicks anymore.

If an LLM can summarize it in one sentence, users never hit your site.

Some suggestions that Ross suggests: 
1. Move resources toward bottom funnel. 
2. Build out pricing pages. 
3. Create detailed comparisons. 
4. Add interactive tools like calculators or templates. 
5. Refresh your case studies.

And if you're going top funnel, make it complex, like use visuals, embed video, add proprietary data. 

Just make it multimodal enough that it can't be compressed into a snippet.

The main takeaway is that search is definitely going away, simply the bar for content is just getting higher.

Done-For-You B2B SaaS Marketing | Founder @ SuperMarketers: Generate More Leads Effortlessly | 1x SaaS Founder & Exit | Book A Call To Learn How 👇

A lot of AI content still seems AI and robotic because people give the model way too much freedom.

AI writes by choosing the most predictable next phrase. 

This pattern creates the same rhythm, the structures, and the same “AI cadence” everyone can spot instantly.... (ie "it's not X, it's Y").

But if you tell the LLM to “sound human”, it is ineffective because the model simply picks more predictable language that fits that request.

So if you want the writing to feel human, you have to break the model’s probability loop.

You can start by forcing a specific voice instead of adjectives, ideally your own! 

Specifically named styles collapse the generic phrase pool and push the model out of autopilot.

Also, add oddly specific details that only a real person would mention.

Humans anchor to moments and specifics, while models default to generalities unless you pin them down.

Another tip is to create a running list of common phrases and structures that models lean on.

The goal is to eliminate the smooth, symmetrical phrasing that gives away AI-generated work.

In short, don’t ask the model to “be human.”

You give it constraints that make predictability impossible.

I’ve been rebuilding my content workflow inside Claude Code, and it’s already changing how I work.

I'm not technical, and kinda regret the initial apprehension about using a "coding" tool....ChatGPT and Claude are definitely sufficient to bridge that gap. 

The screenshot shows part of the system I’m building to turn raw transcripts into structured insight...it can be for a podcast, sales transcript, internal call, etc. 

The folders on the left are not code projects, but just steps in a system that turns raw thinking into structured content.

This is probably how we will all work in the near future, creating small agents to handle a specific part of the process. 

Using this repeatable logic saves time and creates consistency.

For example, one agent takes a podcast transcript and turns it into the things I actually need, like Title, description, time stamps, etc. 

Another agent applies the AEO lens. It looks at what parts of that transcript are extractable, highlights ideas that match the patterns LLMs tend to surface and the pieces that help a concept travel into AI search.

Another agent turns those insights into outlines and content plans. This is near automated and instantaneous, with better quality because the inputs are structured.

In short, Claude Code forces any repeatable task intoa workflow that runs the same way every time.

Still early, but this approach already produces clearer insights and cleaner AEO structure. It also gives me more confidence in the content I deliver for clients.

Google just dropped Gemini 3 and it finally understands the world the way we do.

It reads video, audio, images, movement, and text together as one coherent input.

It goes beyond pattern matching and actually reasons through what is happening in a clip.

You can upload any video and get a true breakdown of what is happening frame by frame.

I tested it by uploading a video of my basketball shot to see if it could analyze motion, timing, pace, and mechanics.

It gave me strengths, weaknesses, corrections, and a plan based only on a few short clips.

But if you want more helpful applications as a marketer: 

👉 You can upload your top ads and ask Gemini why they land, why others flop, and where creative fatigue is creeping in.

👉  You can upload competitor ads to get a precise breakdown of hooks, pacing, messaging gaps, and unique angles they are winning with.

👉  You can record your landing page and ask Gemini to identify friction points, confusing copy, layout issues, and missed opportunities.

👉  You can pair video with analytics and ask which frames or moments likely cause drop-off or bounce.

👉  You can upload sales call recordings and transcripts to surface objections, tone shifts, emotional cues, and coaching moments.

👉  You can upload a product demo from a competitor and get a structured comparison of positioning, missing claims, and angles you should counter.

This is the first time marketers can turn raw video into structured insights without manual review.

This is also a preview of how search and discovery will evolve as multimodal understanding becomes the default.

And yes, I have a long way to go to improve my basketball shot...when asked who my shot resembles, one name was Kevin Martin, historically one of the funkiest shots in league history 😂

I built a small AI agent this week that does something I used to do manually.

It watches what’s being posted on LinkedIn in a niche. 

It captures what gets engagement. 

Then it analyzes the post the way an LLM would. 

Topics, hooks, angle, and how the idea is structured.

The output lives in Airtable, so it’s not just “inspiration”, but actually queryable intelligence.

How it works: 
	•	Pulls posts from specific thought leaders in a topic area, for me that was AEO and AI marketing.
	•	Stores the original post link and basic context.
	•	Adds structured metadata so you can filter and learn patterns fast:
	•	topic cluster
	•	hook type
	•	content format
	•	angle
	•	engagement proxy

This "intelligence hub" is actually a way to build leverage too. 

I’m not a developer. The build is mostly just clarity on inputs and outputs.

The setup for this: 
	•	Apify collects the posts.
	•	Claude Code is the control layer. I describe what I want in natural language, and it generates the logic plus the structure.
	•	Airtable is memory. The hub gets smarter every time new posts land.

One underrated move: 

I asked Claude Code to design the Airtable schema first. It gave me the fields and field types. I pasted it into Airtable and the table was basically done in under a minute.

AI doesn't generate more net new content, but is used to replace: 
	•	manual research
	•	random ideation
	•	low-signal posting

With:
	•	a living dataset of what the market is rewarding
	•	repeatable patterns you can actually see and filter
	•	direction that compounds over time

Oh and also inside the agent, there is a sub-agent for things like:
	•	hook optimization
	•	voice consistency checks
	•	banned phrases and banned structures

That’s how you keep outputs on-brand instead of generic.

WHat other agents would be helpful to build out for marketers?

LLMs do not read your article in its entirety.

It slices it into small chunks and only keep the chunks that express a single idea with absolute clarity.

If a chunk is scattered, the model drops it. and conversely, if a chunk is clear, it becomes part of an answer.

Here are five simple ways to create content chunks that are valued by LLMs: 

- A definition.
- A set of steps.
- A comparison.
- A direct question and answer.
- A real example.

These formats survive chunking, get reused in answers, and scale your ideas without extra effort.

Long rambling paragraphs that feel smart but break into unusable pieces....this was my (cringey) writing style in college, seemed to make sense but is not valuable then, or now. 

Write in a way that makes your ideas easily extractable, for LLMs and your readers. 

One idea per chunk.
One chunk per format.
Five formats that consistently show up in AI answers.

That is the blueprint.

Google just made a quiet but important change to AI search.

When you search something in AI Mode, Google now includes WHY certain sources are relevant.... before it shows the links.

This is important and relevant because AI search is no longer just about ranking, or being mentioned or cited, but being explainable is now important. 

Google’s AI is literally telling users “Here’s why I trust these sources.”

Content that wins now is:
 • clearly scoped
 • structurally organized
 • explicit about the problem it solves
 • easy for an AI to summarize and justify

And generic content is harder to explain. Death to AI slop 😅 

So when creating content: 
- Write content AI can summarize in one sentence.
- Use definitions, steps, examples, and FAQs.
- Build content systems, not one-off posts.

Structure matters a lot for visibility in AI search, and this is one more step in that direction.

Jon Nordmark

Iterate.ai

Iterate.ai is one of only three startups named to CRN’s 2025 list of the 20 Hottest AI Software Companies. 

Iterate’s primary focus: delivering Private AI for the Enterprise. 

As a full-stack AI company, Iterate helps global brands move faster, safer, and cheaper. Its two platforms — Interplay for developers and Generate for business users — let enterprises build and run AI agents across public clouds, private clouds, air-gapped appliances, AI PCs, and far-edge devices, even in places without internet or electricity. Interplay’s runtime has cut LLM inference costs by up to 95% while increasing speed. Generate extends these capabilities with secure, no-code tools for knowledge work and automation.

The company also built AgentOne, an enterprise-grade agentic coding product. Designed as an extension of Interplay and Visual Studio, it comes with an OWASP WSTG security test, 18 independent security scans, static security code analysis, architecture mapping, and unit testing built in. AgentOne enables teams to design, test, and launch applications up to 80X faster.

Trusted by global leaders at companies like IBM, Ulta, e.l.f., MUFG (Japan’s largest bank), Great Eastern Life, FUJIFILM Singapore/APAC, Intel, Qualcomm, NVIDIA, and Pure Storage, Iterate unlocks AI innovation at each layer of the generative AI tech stack (application, model, and infrastructure layers).

Co-founder, CEO

Colorado Technology Association

Board Member

eBags.com (acquired by Samsonite)

After selling 29 million bags (worth $1.65 billion), eBags was acquired by Samsonite in 2017. Annual revenues were $164 million, exceeding $1.5 million per employee. eBags brand (private label) was 25% of sales. About 70% of products were drop shipped ... fueling a 120 day negative cash conversion cycle and growth that was funded by customer credit cards. 

During my final 8 years as CEO (2001 to 2008), we grew profitably at a 34% CAGR. Customers wrote more than 1.4 million product reviews making eBags a trusted retailer. We operated TUMI.com and CaseLogic.com, in addition to eBags, in the U.S., U.K., and Japan. We also built 6pm.com, acquired by Zappos and owned today by Amazon.

eBags' first profitable month was August 2001, just ahead of the 9/11 World Trade Center tragedy, which greatly dampened luggage sales for more than 6 months. Our first profitable quarter was in 2002 after growing through the dot-bomb recession and the months following the 9/11 terrorist attack. 

We raised $30 million prior to the dot-bomb and took great pride in reaching profitability, growing 34% per year by continually reinventing our marketing efforts, and never doing a down round.

Awards: 
@ Website of the Year from the Direct Marketing Association in 2001, 2002, 2004, 2006, 2007 and 2008; 
@ Entrepreneur of the Year from Ernst & Young, 2003, Rocky Mountain Region; 
@ Company of the Year in 2003 from Colorado Business (magazine);
@ Top customer satisfaction scores (Circle of Excellence) from BizRate twice (based on shopper scores of >50,000 websites).

Co-founder, CEO and Chair from 1998 to 2008; Chairman from 2008 until acquired in 2017.

Co-founder, CEO, Chairman

Runa Inc. (acquired by Staples)

Angel investor in A/B testing optimizer Clearhead (acquired 2017, Accenture), BazaarVoice (pre-IPO, 2012), Swiftpage/ACT! (acquired 2018), RedTricycle (acquired 2020), Verblio.com (exit 2012), Yottaa (exit 2022), Platform9, AX3.

4x LP in LeadEdgeCapital.com ... invests in growth-stage software, internet, and consumer pre-IPO businesses like Spotify, Uber, Bla Bla Car, and Alibaba. Top performing PE in 2022 according to Pitchbook.

Advisory board chairperson for Runa (acquired 2017, Staples), shareholder and advisor to TurnTo.com (bootstrapped to 80 employees, exit 2022).

Iterate became a shareholder in Attentive ($6b unicorn) and BlueCore ($1b unicorn, exit in 2021 secondary offering). 

More info: https://signal.nfx.com/investors/jon-nordmark

Between 2009 and 2013, worked with dozens of emerging tech startups. Many from TechStars (Boulder), Founder Institute (Palo Alto) and an Eastern European Accelerator where I was a Board Member for 3 years.

Advisor; Investor ... many startups

6pm.com (acquired by Zappos)

Deming Center for Entrepreneurship

Samsonite

Product positioning, pricing, and development for 21 distribution channels (Premium, Department Stores, Military, etc). Marketing, collateral development, sales force training. About 1,200 products continually in circulation and being replaced. Worked primarily in the U.S., but also in Hong Kong, China, Europe, Mexico and Argentina.

Senior Director of Product Management, N. & S. America

* E&Y Entrepreneur of Year, Rcky Mtn Rgn
* Colorado CEO of the Year, CTA
* LinkedIn Top Voice for 20+ tech articles
* Lifetime Achievement Award, CTA


Iterate awards:
☆ 20 Hottest AI Software Companies, Channel Reseller News (CRN) 2025
☆ Best Workplaces for Innovators 2024: AI and Robotics, Iterate.ai, Fast Company 2024
☆ KW World AI 100 in 2024, 2025
☆ Technology of the Year Awards: AI and machine learning: Models, Iterate.ai InfoWorld 2024
☆ Best in Business for AI and Data, Iterate.ai, Inc. Magazine 2023, 2024
☆ Company of the Year, Iterate.ai, Colorado Technology Association, 2024
☆ CEO of the Year, Iterate.ai, Colorado Technology Association, 2025
☆ Lifetime Achvmt Award | CO Tech Assoc.

eBags awards:
☆ Dirct Mktng Assoc Website of Yr, eBags: 2001, 02, 04, 06, 07, 08

My current mission is to empower action-oriented executives with AI.

My company, Iterate.ai, built Interplay, a patented low-code AI platform that enables enterprises to rapidly build and deploy AI solutions across the entire Generative AI stack—from infrastructure to models to applications—even on resource-constrained edge devices and securely behind firewalls. Trusted by tech leaders like IBM, Intel, and PureSoft, as well as enterprises like Ulta Beauty and MUFG (Japan’s largest bank), Iterate unlocks AI opportunities at every level. From custom AI development to ready-to-use applications like Generate—all built on Interplay—Iterate scales through a recurring software licensing model that drives high gross margins. With its ability to build and deploy networks of AI agents, Iterate.ai is uniquely positioned to lead the next wave of Agentic AI.

Iterate.ai:
Founded:  2013			
AI Focus:  Iterate.ai — February 2015
Funding:  $3 million up to 2023, $7 million more in late 2024-5 (primarily angel investors)
    -- We are very capital-efficient and inventive
Employees and Contractors:  109, mostly engineers (as of Mar 2025)
Primary Offices:  San Jose, CA and Denver, CO

As the co-founder and CEO of Iterate.ai, my team of innovators and creators help enterprises accelerate AI adoption and digital transformation. Thanks to my 25+ years of experience in online retail and e-commerce, I have a deep understanding of the needs and challenges of customers and businesses in the digital age. Since 2015, we've been solving challenges by applying AI.

543% growth over the past five years. Grew to >$10 million in revenue with an $82,000 annual marketing budget and no salesperson.

Nvidia (AI Accelerated), Google Cloud, AWS are tech partners, as are others.

CDOs, CMOs, and CIOs depend on Iterate to speed up Digital Transformation in a modern way. 

90% of Iterate's revenues are from our patented low-code platform, Interplay.  Enterprises use Interplay to accelerate AI adoption. Action-oriented executives choose Interplay because it facilitates up to 17x faster prototyping of digital, AI, and IoT solutions. Those apps scale straight into production. Interplay even processes AI on Edge Computers in 3,000+ locations ... from convenience stores to oil change facilities ... in >10 countries (Estonia to Sweden and Denmark to USA). 

Interplay's engineers have written >500 pre-built software components; 43 of the 500 are AI enablers like Random Forest Classifiers. Interplay also makes >3,500 reusable components available from the dev community.

Interplay is supported by a second proprietary product named Signals. Signals helps executives make informed decisions, build digital solution roadmaps and architectures. It uses AI to monitor 16 million startups, >100 million patents, and 100s of emerging trends. 

Companies from diverse industries impacted by AI and IoT (banking, insurance, retail, entertainment, education, telecom, luxury, CPG) partner with Iterate.ai.

𝗣𝘂𝗯𝗹𝗶𝗰 𝗔𝗜 may create the 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗮𝗰𝗰𝗶𝗱𝗲𝗻𝘁𝗮𝗹 𝗱𝗮𝘁𝗮 𝗹𝗲𝗮𝗸 𝗶𝗻 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗵𝗶𝘀𝘁𝗼𝗿𝘆.

Public AI includes ChatGPT, Gemini, Grok, Anthropic, DeepSeek, Perplexity — any chat system that runs on giant 𝘀𝗵𝗮𝗿𝗲𝗱 𝗚𝗣𝗨 𝗳𝗮𝗿𝗺𝘀.

And while those public models do amazing things, 
we need to talk about something most leaders underestimate.

— 𝗣𝘂𝗯𝗹𝗶𝗰 𝗔𝗜 𝗶𝘀𝗻’𝘁 𝗮 𝘃𝗮𝘂𝗹𝘁.
— 𝗜𝘁’𝘀 𝗮 𝘃𝗮𝗰𝘂𝘂𝗺.

People are pasting their identities, internal documents, and even corporate secrets into systems designed for scale — 𝗻𝗼𝘁 𝗽𝗿𝗶𝘃𝗮𝗰𝘆.

It feels private.
— But —
It isn’t.

Analogy: 
Using Public AI is like whispering confidential strategy into a megaphone because you thought it was turned off.

Now pause for a moment and watch the video.

The colored lines represent 𝗺𝗶𝗹𝗹𝗶𝗼𝗻𝘀 𝗼𝗳 𝗽𝗿𝗼𝗺𝗽𝘁𝘀 from 𝗺𝗶𝗹𝗹𝗶𝗼𝗻𝘀 𝗼𝗳 𝗽𝗲𝗼𝗽𝗹𝗲 working at 𝘁𝗵𝗼𝘂𝘀𝗮𝗻𝗱𝘀 𝗼𝗳 𝗰𝗼𝗺𝗽𝗮𝗻𝗶𝗲𝘀 — running across:

— 𝘀𝗵𝗮𝗿𝗲𝗱 𝗵𝗮𝗿𝗱𝘄𝗮𝗿𝗲 (1,000s of NVIDIA GPUs powering Public AI)
— 𝘀𝗵𝗮𝗿𝗲𝗱 𝗺𝗼𝗱𝗲𝗹𝘀 (like ChatGPT’s GPT-5)
— 𝘀𝗵𝗮𝗿𝗲𝗱 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲

"𝑺𝒉𝒂𝒓𝒆𝒅" is operative word... 
   the word to pay attention to.

T.H.A.T. is Public AI: 
powerful, massive, centralized… but fundamentally risky.

Especially for CISOs, Boards, and CEOs responsible for safeguarding PII, HIPAA-sensitive, and financial data.

When your data enters a Public LLM, it moves across the world:
— It gets logged.
— It gets cached.
— It gets stored.
— And sometimes, it gets trained on.

Even when vendors like OpenAI (ChatGPT), Google (Gemini), DeepSeek, Anthropic, and Perplexity say they don’t train on your data, 𝗼𝘁𝗵𝗲𝗿 𝗿𝗶𝘀𝗸𝘀 𝗿𝗲𝗺𝗮𝗶𝗻:

— logging
— retention
— global routing
— caching
— prompt injection
— model leakage
— subpoena exposure

Training isn’t the only danger. 
It’s just one of many.

You may think you “deleted” something… but think again.

That’s why this series exists:
To break down the overlooked risks of Public AI — and highlight the safer path.

That safer path is Private AI:
Models that run behind your firewall, on your hardware, with your controls.

You may think you “deleted” something, but think again.

Tomorrow we begin the list.

𝗣𝗿𝗶𝘃𝗮𝘁𝗲 𝗔𝗜 🔒 𝗸𝗲𝗲𝗽𝘀 𝘆𝗼𝘂𝗿 𝗱𝗮𝘁𝗮 𝗯𝗲𝗵𝗶𝗻𝗱 𝘆𝗼𝘂𝗿 𝗳𝗶𝗿𝗲𝘄𝗮𝗹𝗹 — 𝗻𝗼𝘁 𝗶𝗻 𝘀𝗼𝗺𝗲𝗼𝗻𝗲 𝗲𝗹𝘀𝗲’𝘀 𝗹𝗼𝗴𝘀.

— I’ve added a simple explainer in the comments.

(This post is part of my 27-part Public AI Risk Series.)

#PrivateAI #EnterpriseAI #CyberSecurity #BoardDirectors #AICompliance

Co-founder, CEO @ Iterate.ai - private AI || co-founder, CEO @ eBags - $1.6B of products sold, then acquired

The Economist just revealed a surprising fact: 
AI adoption in big companies is falling. 
-- Not rising. 
-- Not plateauing. 
It's falling.

Here’s why it’s happening — and how to fix it.

#AI #EnterpriseAI #PrivateAI

Two days ago, The Economist published a sharp and underreported signal about enterprise AI adoption. Using new Census Bureau data, they found: AI use 

Open article: The Economist's alarm: Enterprise AI adoption is stalling—especially at big firms. Here’s why. And how to fix it. by Jon Nordmark • 9 min read

The Economist's alarm: Enterprise AI adoption is stalling—especially at big firms. Here’s why. And how to fix it.

AI now writes code faster than any enterprise can review it. 

That gap is becoming one of the biggest security risks in modern software.

I just shared why we built AgentOne 
— and how real-time security must evolve to match AI’s new speed.

AI now generates code at a velocity no enterprise engineering team has ever faced — orders of magnitude beyond human output. Human developers used to 

Open article: Enterprise Vibe Coding Just Hit a Turning Point by Jon Nordmark • 3 min read

Enterprise Vibe Coding Just Hit a Turning Point

𝗦𝗵𝗮𝗱𝗼𝘄 𝗔𝗜 𝗜𝘀 𝘁𝗵𝗲 𝗤𝘂𝗶𝗲𝘁𝗲𝘀𝘁 𝗖𝗼𝗿𝗽𝗼𝗿𝗮𝘁𝗲 𝗟𝗲𝗮𝗸 𝗛𝗮𝗽𝗽𝗲𝗻𝗶𝗻𝗴 𝗧𝗼𝗱𝗮𝘆

Employees don’t mean to leak company secrets.
 They just want to get work done faster.

So they paste:
 — Board decks
 — Client lists
 — Forecasts
 — Emails
 — Pricing

Straight into public AI tools.

And here’s where things quietly go off the rails.
A recent study found that 𝟵𝟯% 𝗼𝗳 𝗲𝗺𝗽𝗹𝗼𝘆𝗲𝗲𝘀 𝗮𝗱𝗺𝗶𝘁 𝘁𝗼 𝘂𝘀𝗶𝗻𝗴 𝘂𝗻𝗮𝘂𝘁𝗵𝗼𝗿𝗶𝘇𝗲𝗱 𝗔𝗜 𝘁𝗼𝗼𝗹𝘀 𝗳𝗼𝗿 𝘄𝗼𝗿𝗸.

 Even more surprising: generative AI has already become one of the top channels for accidental corporate data exfiltration.

This isn’t malicious.
 — It’s human nature.

People reach for the fastest tool available.
 And right now, the fastest tool is public AI.

Here’s another subtle problem: 
The chat interface 𝗲𝗻𝗰𝗼𝘂𝗿𝗮𝗴𝗲𝘀 𝗼𝘃𝗲𝗿𝘀𝗵𝗮𝗿𝗶𝗻𝗴.
 —  It’s designed to be friendly.
 —  It feels private.
 —  No one is watching over your shoulder.
 And humans always share more when something “feels” safe.

But here’s reality:
𝗣𝘂𝗯𝗹𝗶𝗰 𝗔𝗜 𝗶𝘀𝗻’𝘁 𝗮 𝘃𝗮𝘂𝗹𝘁.
 𝗜𝘁’𝘀 𝗮 𝘃𝗮𝗰𝘂𝘂𝗺.

It doesn’t hold your data securely.
 It absorbs it—into logs, retention layers, training pipelines, and systems you will never see.

And that’s the trap.
𝗨𝘀𝗶𝗻𝗴 𝗮 𝗽𝘂𝗯𝗹𝗶𝗰 𝗔𝗜 𝗰𝗵𝗮𝘁𝗯𝗼𝘅 𝗶𝘀 𝗹𝗶𝗸𝗲 𝗼𝗽𝗲𝗻𝗶𝗻𝗴 𝘂𝗽 𝘁𝗼 𝗮 𝘁𝗵𝗲𝗿𝗮𝗽𝗶𝘀𝘁… 𝗼𝗻𝗹𝘆 𝘁𝗼 𝗱𝗶𝘀𝗰𝗼𝘃𝗲𝗿 𝘁𝗵𝗲 𝘁𝗵𝗲𝗿𝗮𝗽𝗶𝘀𝘁 𝗶𝘀 𝘄𝗲𝗮𝗿𝗶𝗻𝗴 𝗮 𝘄𝗶𝗿𝗲.

That’s Shadow AI.

𝗦𝗵𝗮𝗱𝗼𝘄 𝗔𝗜 (𝘀𝗶𝗺𝗽𝗹𝗲 𝗱𝗲𝗳𝗶𝗻𝗶𝘁𝗶𝗼𝗻):
Shadow AI is when employees use public AI tools without approval or security controls—quietly leaking internal data into systems their company doesn’t own.

The model may retain your data.
 The logs may store it.
 Engineers may see it.
 Other users may benefit from it indirectly.
 And the vendor may train future models on it.

𝗢𝗻𝗰𝗲 𝘆𝗼𝘂𝗿 𝗰𝗼𝗺𝗽𝗮𝗻𝘆’𝘀 𝗜𝗣 𝗲𝗻𝘁𝗲𝗿𝘀 𝗮 𝗽𝘂𝗯𝗹𝗶𝗰 𝗟𝗟𝗠, 𝗶𝘁 𝗺𝗮𝘆 𝗻𝗲𝘃𝗲𝗿 𝗰𝗼𝗺𝗲 𝗯𝗮𝗰𝗸.

𝗦𝗵𝗮𝗱𝗼𝘄 𝗔𝗜 𝗔𝘂𝗱𝗶𝘁 (𝘶𝘯𝘥𝘦𝘳 𝘢𝘵𝘵𝘰𝘳𝘯𝘦𝘺–𝘤𝘭𝘪𝘦𝘯𝘵 𝘱𝘳𝘪𝘷𝘪𝘭𝘦𝘨𝘦)
Most organizations don’t know how employees are using public LLMs.
To fix this, Iterate can help you get a confidential Shadow AI Audit—under attorney–client privilege—to assess exposure and plan safer, governed alternatives.

If you’d like to learn more, just let me know.

𝙋𝙧𝙞𝙫𝙖𝙩𝙚 𝘼𝙄 𝙛𝙞𝙭𝙚𝙨 𝙩𝙝𝙞𝙨 𝙖𝙩 𝙩𝙝𝙚 𝙨𝙤𝙪𝙧𝙘𝙚.
It gives employees a powerful internal system—behind your firewall—so they don’t need public AI tools to be productive.

🌐 𝗣𝘂𝗯𝗹𝗶𝗰 𝗔𝗜 𝗹𝗲𝗮𝗿𝗻𝘀 𝗳𝗿𝗼𝗺 𝘆𝗼𝘂𝗿 𝗱𝗮𝘁𝗮. 𝗣𝗿𝗶𝘃𝗮𝘁𝗲 𝗔𝗜 𝗽𝗿𝗼𝘁𝗲𝗰𝘁𝘀 𝗶𝘁.
 I’ve added a simple explainer in the comments.

(𝘗𝘰𝘴𝘵 3 𝘰𝘧 ~27 𝘪𝘯 𝘮𝘺 𝘗𝘶𝘣𝘭𝘪𝘤 𝘈𝘐 𝘙𝘪𝘴𝘬 𝘚𝘦𝘳𝘪𝘦𝘴)
#AI #ShadowAI #Cybersecurity #DataSecurity #PrivateAI

taplio

The Battle

🌷 Gen Furukawa

Jon Nordmark

Followers

Posts

Engagements

Likes

Performance Score🌷 Gen Furukawa's LinkedIn Performance

Performance ScoreJon Nordmark's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

🌷 Gen Furukawa

Jon Nordmark

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}🌷 Gen Furukawa's LinkedIn Performance

Performance ScoreJon Nordmark's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance Score🌷 Gen Furukawa's LinkedIn Performance