Token Factory Suppliers
Clay Norris@claynorris
+7.0%
since Oct 1
The idea
I believe AI agents are stuck without two things they can't generate themselves: licensed proprietary data and verified identity, and the deals to supply both are landing within the next couple of years. So I built this around three groups: companies already signing AI licensing agreements, the data monopolies holding the pricing power for when their turn comes, and the identity and trust rails agents will need to verify who they're acting for.
What it owns
- Before an agent can move money or sign anything, it must verify who it acts for, and these companies sell that verification infrastructure. · Okta Inc.It builds the layer that verifies which entity an AI agent is acting for, the trust infrastructure every agent transaction requires., MastercardIt puts verified credentials on transactions so agents can prove they are authorized to spend money, owning the trust layer for payments., SailPointIt controls which machine identities can access which data inside companies, a problem that grows more important with each agent deployed., Clear SecureIt owns a biometric network that ties a real person to a digital identity, the root identity token agents act on behalf of., Mitek SystemsIt is the pure play on document and identity verification, the checks standing between agents and fraud that become critical at scale.26%
- Their data cannot be scraped or recreated, forcing AI builders to eventually pay for access to proprietary information. · Moody's CorporationIts credit ratings are un-scrapable data that agents cannot produce themselves, gaining pricing power as AI adoption expands., Verisk AnalyticsIt holds the insurance industry's shared data pool, the expertise no model builder can recreate that agencies must access., S&P GlobalIt owns proprietary ratings and benchmark data that financial market participants cannot easily replicate or access without paying licensing fees., FactSet Research SystemsIt delivers cleaned financial data into workflows and is building AI products on top of its own data, becoming more valuable as AI scales., Intercontinental ExchangeIt sells market data feeds that any financial agent needs to act, already charging for access in a model that works today.26%
- These companies have already signed paid licensing deals for their content and data with AI builders, demonstrating the revenue model this take depends on. · Reddit Inc.It has signed licensing deals with AI model builders for conversational data no one else possesses, proving this take is already happening., Thomson ReutersIt owns decades of legal and professional data that cannot be scraped, exactly the expertise tokens agents need for professional-grade work., News CorporationIt signed one of the first major AI content licensing deals, establishing a template for the wave of licensing agreements this take expects., New York Times Co.Its un-scrapable archives plus legal leverage position it to command rich licensing deals from model builders rather than grant free access.24%
- Credit files and scores represent the deepest financial records of individuals, exactly what agents making money decisions will need to license. · EquifaxIt owns income and employment records agents cannot obtain elsewhere, the personalization data that becomes more valuable as agent lending grows., Fair Isaac CorporationIt owns the score every credit decision relies on, a toll with no substitute that survives only if agents pay it rather than replace it., TransUnionIts consumer credit files are context tokens with no substitute that any agent acting on someone's finances must access.16%
- Cash8%