PLUS: Monetising your AI training data, SAP's $300M health move, and IBM's 42% productivity forecast

Afternoon all,

"Whosoever commands the trade of the world commands the riches of the world, and consequently the world itself" - Sir Walter Raleigh

First it was the seas, then it was Gold, then Oil, and now it is Data. Today and in tomorrow's economy if you control data, you can control the riches of the world and by extension the world itself. Cloudflare believe that and are putting their money behind that belief. Their $1B acquisition of Human Native AI positions the company at the center of AI's most contentious debate: who profits when machines learn from human-created content?

A new data licensing framework could flip the script for creators seeking equitable compensation. However, will corporate data gatekeepers ultimately determine what 'fair use' means in the age of large language models?

Today's dots:

• Cloudflare's ethical AI training marketplace
• SAP & Fresenius' €300M healthcare AI play
• Runway's production-ready Gen-4.5 video
• IBM forecasts 42% AI productivity leap

Cloudflare bets $1B+ to make AI training fair

Here's the thing: Cloudflare just acquired London-based Human Native AI in a deal reportedly worth over $1 billion, aiming to build ethical data pipelines that compensate creators when their work gets used for AI training.

Let's unpack that:

• Human Native specialises in turning unorganised content into structured, licensable datasets, consequently giving publishers and creators control over how their work feeds AI models
• Cloudflare's introducing new tools like Pay Per Crawl and AI Crawl Control that let creators charge per AI training visit and not just block automated scraping
• This comes as tensions rise between AI companies and publishers over unauthorised data use. Think lawsuits like The New York Times v. OpenAI but with proactive licensing solutions
• Cloudflare's building a payment infrastructure (x402) that could enable automatic micropayments every time AI systems train on licensed material
• Their new AI Index lets developers access fresh structured content directly -reducing messy web scraping while tracking usage transparently

If you remember nothing else: This acquisition signals a power shift towards content creators in the AI training economy. This includes brands and retailers. In this new era of agentic commerce business models will be ripped up and reformed. Those who own valuable training data could soon monetise it as effortlessly as running display ads.

SAP & Fresenius Bet €300M on Sovereign Healthcare AI

Here's the thing: SAP and healthcare giant Fresenius just committed €300M to build Europe's first fully compliant healthcare AI platform (original announcement), combining SAP's Business Technology Platform with Azure AI to tackle sensitive patient data.

Let's unpack that:

• This creates a sovereign data fortress where European hospitals can use AI without sending patient records overseas - crucial for GDPR compliance in the age of diagnostic chatbots and predictive analytics
• The platform uses open standards like HL7 FHIR to connect fragmented hospital systems - imagine your local clinic's records instantly syncing with specialist AI diagnostic tools at university hospitals
• SAP's bringing its enterprise-grade SAP Business AI tools to healthcare for the first time, letting hospital admins automate bed allocation while clinicians get AI-assisted treatment recommendations
• Their 'AnyEMR' strategy could finally fix healthcare's interoperability nightmare - no more need for sending analog records between clinics that use different electronic systems
• It's a tactical play in the global AI race: Europe's answering US/China's healthcare AI efforts with a privacy-first alternative that protects sensitive health biometrics (EU's digital health strategy)

If you remember nothing else: This isn't just another healthcare IT project - it's Europe planting its flag in the ethical AI landscape. Europe was conspicuously absent from ChatGPT Health due to privacy issues. European hospitals getting this right could set global standards for how we handle AI in life-or-death scenarios while keeping patient agency intact.

Runway's Gen-4.5 Makes Pro-Grade Video Production Accessible

Here's the thing: Have you ever had that delusional confidence that you could make a Citizen Kane level movie but the only thing stopping you was money and the Hollywood machine? Well now there are no more excuses. Runway's new Gen-4.5 model tops the Artificial Analysis video benchmark, delivering studio-quality motion control and physics simulation into the hands of marketers and indie creators alike.

Let's unpack that:

• Motion precision lets you maintain character consistency across shots - no more floaty avatars or warped objects between frames
• Physics engine simulates realistic water flow, object collisions and material properties that previously required expensive VFX packages
• Handles everything from photoreal product shots to stylised animation with the same prompt adherence - perfect for agile ad campaigns
• Runs on NVIDIA Blackwell architecture, delivering compute efficiency that keeps generation costs predictable at scale
• Already adopted by enterprise clients like D & G, Target and Wieden Kennedy for rapid prototyping of product visuals

If you remember nothing else: This isn't just about fancier AI clips. Gen-4.5 removes cost barriers that previously kept small teams from competing with studio productions. Your next brand film might cost 90% less than last year's version.

AI Prepares to Deliver 42% Productivity Surge by 2030

Here's the thing: AI will drive a 42% productivity boost by 2030 according to IBM's landmark study of 2,100 executives - with 79% expecting AI-driven revenue growth this decade (full report).

Let's unpack that:

• Manufacturing and healthcare lead AI adoption, with carmakers already using automation savings to develop AI-powered vehicles that predict maintenance needs and update features over-the-air
• 80% of IT services firms now reinvest efficiency gains into outcome-based offerings - shifting from billing hours to selling measurable business results
• Aerospace companies are prioritising custom AI models for secure operations - with 79% already defining their 2030 AI architecture needs versus just 24% in other sectors
• Healthcare organisations use AI for back-office admin, freeing doctors for complex care decisions - leading to new roles that combine clinical oversight with AI supervision
• Quantum computing could trigger the next disruption wave - banking leaders predict breakthrough applications in fraud detection despite only 27% having active quantum plans

If you remember nothing else: Companies aren't just chasing efficiency - they're building AI into their business DNA. The winners will transform entire industries through continuous reinvention, not just cost-cutting.

The Shortlist

IBM launched Enterprise Advantage, a $58M service helping companies deploy agentic AI across workflows while maintaining governance controls. The offering builds on IBM's internal AI tools that boosted consultant productivity by 50%.

VCs bet $1.7B on AI security startups in 2025 as agentic systems create novel risks – Ballistic Ventures' portfolio company WitnessAI grew 500% YoY solving 'shadow AI' threats where autonomous agents access unauthorized systems.

Philippines emerged as an AI-human hybrid hub for global finance, with JPMorgan and Citi employing 250K workers alongside AI systems that handle 40% of fraud detection and loan processing at 1/3 Western operational costs.