In partnership with
Hey everyone,
Remember when researchers said we'd run out of training data by 2026?
It's 2026. They were right.
Not of content, but of training data worth scraping. Every AI lab hit the same wall in 2024. They fed Wikipedia, Reddit, news sites, and academic papers into their models, and now there's nothing left.
OpenAI is fighting publishers in court. Reddit locked its API. X now charges $5,000/month for Pro access, jumping to $42,000/month for Enterprise data. The free training data era? Dead.
But here's the twist: some companies saw this coming and built the solution before anyone else noticed.
📊 Synthetic Data by the Numbers
Real Data Still Matters
Synthetic data trains models. Real customer data runs your business. Attio keeps it clean, organized, and accessible without the enterprise bloat.
→ Try Attio here ⤵
Introducing the first AI-native CRM
Connect your email, and you’ll instantly get a CRM with enriched customer insights and a platform that grows with your business.
With AI at the core, Attio lets you:
Prospect and route leads with research agents
Get real-time insights during customer calls
Build powerful automations for your complex workflows
Join industry leaders like Granola, Taskrabbit, Flatfile and more.
Why It's Exploding Right Now
Synthetic data means AI-generated datasets that look and behave like real data but aren't. And it just went from experiment to necessity.
Banks simulate fraud that happens once in a million transactions. Hospitals train diagnostic AI on synthetic patient records without violating GDPR. Self-driving cars test crash scenarios they can't ethically create.
Over 60% of AI applications already use some form of synthetic or augmented data. Financial services teams report 40-60% faster development timelines. The economics are too good to ignore.
💡 Builder Spotlight: Who's Winning
Machine learning for fraud detection and risk modeling
Enterprise synthetic data generation platform
Deutsche Bank is both investor and customer
42 startups competing with $763M in total funding
The Part Nobody Tells You
Train a model only on synthetic data and it collapses. Outputs get worse each generation until the whole thing breaks.
But mix real and synthetic correctly? Performance improves.
The ratio is everything. Too much fake data = model degradation. Too little = you're not solving the scarcity problem. Companies figuring out this balance are winning their sectors.
One innovation director put it bluntly: "The appeal is obvious. Speed, cost savings, and the ability to simulate those elusive consumer segments".
⚡ Real World Impact
- Healthcare: Train diagnostics AI without patient privacy risks
- Finance: Stress test portfolios and simulate market crashes
- Autonomous Systems: Test rare scenarios like pedestrian accidents
- Fraud Detection: Generate edge cases that barely exist in real data
Where the Money Is
Most builders focus on models. Almost nobody's building the data layer.
The real opportunity? Vertical-specific generation pipelines for regulated industries:
Healthcare diagnostics
Financial stress testing
Fraud simulation engines
Autonomous vehicle edge cases
These sectors have deep pockets, strict regulations, and urgent need. They'll pay to eliminate privacy risk and cut development from quarters to hours.
The Opportunity
We ran out of free oil. Someone has to build the factory.
The market went from $765M in 2025 to a projected $6.47B by 2032. That's 35.67% annual growth in infrastructure most people don't know exists.
The companies building generation infrastructure today are positioning themselves as the refineries in a data-constrained world. Most builders still don't see it.
Stay sharp,
Better Every Day
📬 Building something in synthetic data? Hit reply. I'm tracking tools and approaches for a future breakdown.
The Future of Shopping? AI + Actual Humans.
AI has changed how consumers shop, but people still drive decisions. Levanta’s research shows affiliate and creator content continues to influence conversions, plus it now shapes the product recommendations AI delivers. Affiliate marketing isn’t being replaced by AI, it’s being amplified.
Connect your teams and data on one CRM.
Help your teams work better together with shared inboxes, deal boards, customer insights, and more. All your data at your fingertips, streamlined by AI. Start free today.
