AI is easy to fake, Theranos whistleblower Tyler Shultz tells SF Standard
THE SO WHAT
When a Theranos whistleblower says AI is hard to vet and lightly regulated, that’s a warning that “AI” claims will follow the same hype-fraud pattern unless investors change their process. If you’re diligencing AI startups, shift from demo-driven evaluation to code, data, and customer-reference audits this quarter.
READ THE SOURCE
MORE FROM THE WIRE
Startups & VentureOpenAI reportedly lowers funding target for new round
A lowered funding target from OpenAI is a reminder that even top labs are normalizing capital intensity and growth expectations—cheap, infinite AI money is not the base case. If you're building on these models, pressure‑test your own runway and pricing against a world where upstream vendors prioritize margin and unit economics over pure scale.
Anthropic's IPO push includes a big bet on life sciences
Anthropic is telling public markets that Claude is a drug discovery engine — not just a general assistant. If you’re in biotech, assume investors will start asking why your stack isn’t wired into these models and be ready to separate real acceleration from marketing claims.
Startups & VentureSoftBank-Backed OPay Halts IPO to Respond to Regulatory Inquiry
A last-minute regulatory inquiry freezing OPay’s NYSE debut shows how fragile cross-border fintech listings still are. If you’re scaling payments in emerging markets, bake regulatory scenario planning into your capital strategy — including delayed IPO windows and alternative liquidity paths.
Startups & VentureSources: neoclouds CoreWeave, Nebius, and Nscale have effectively been shut out of the bond market in recent weeks as lenders and investors demand tougher terms
Cheap debt is no longer a given for GPU neoclouds—Wall Street is now pricing AI infra like project finance, not growth equity. If your roadmap assumes easy burst capacity from second-tier clouds, model in tighter credit, slower buildouts, and more conservative contract terms.