Human-Centered AI
The design principle — associated with Stanford's Institute for Human-Centered AI (HAI) — that AI systems should be built around human needs, values and well-being, not just technical capability. AI-Human Systems is the organisational doctrine that operationalises that principle inside a real company; the two are related, not the same thing.
A design principle, most closely associated with Stanford's Institute for Human-Centered Artificial Intelligence (Stanford HAI), holding that AI systems should be built around human needs, values and well-being — not just around what the underlying model is technically capable of doing.
It's a standard for the artefact: the AI product or system itself. It doesn't specify how the organisation building or deploying that artefact has to be structured for the standard to hold once the system ships. That's a separate, organisational question — the one AI-Human Systems answers.
Context & Strategy
Where it comes from
Stanford HAI was co-founded in 2019 by Fei-Fei Li and John Etchemendy, on the premise that AI development needed an interdisciplinary anchor — computer science combined with psychology, ethics and design — rather than being driven by technical capability alone. Human-centered AI, in that framing, means systems judged not just on accuracy but on whether they end up trustworthy, inclusive, and aligned with the goals of the people and society using them.
Relationship to AI-Human Systems
Human-Centered AI says what a good AI system should be like. AI-Human Systems says what the company around that system has to look like for the design to actually survive a real deployment — who owns the decision when the human-centered design and a throughput target conflict, what authority the people expected to exercise judgment actually have. Treat them as the same term and you lose the distinction that makes either one useful: one is a standard for the artefact, the other is the organisational condition that lets the standard hold. See What Is Human-Centered AI? for the full argument.
Frequently Asked Questions
No. Human-Centered AI is a design philosophy for AI systems. AI-Human Systems is a broader organisational doctrine about architecting strategy, technology and people as one system — AI is one instance of what it applies to, not the whole of it.
It's most closely associated with Stanford's Institute for Human-Centered Artificial Intelligence (Stanford HAI), co-founded by Fei-Fei Li and John Etchemendy in 2019.