South Korea bets on AI adoption with AI Basic Act and eldercare robots

South Korea has decided public enthusiasm is a policy input. A majority of its citizens use AI daily and report more excitement than worry, and the government has built a national strategy on top of that sentiment — one that explicitly favours development over safety.

What South Korea actually did

The evidence is a coherent stack, not a single announcement:

• A national AI strategy backed by the country's semiconductor incumbents, Samsung and SK Hynix.
• AI textbooks deployed into education.
• Eldercare robots aimed at a demographic problem South Korea already faces.
• The AI Basic Act, a law framed to promote AI development rather than constrain it.

The consistent thread is demand-side confidence. Where much of Europe treats citizen unease as the thing to manage, South Korea is treating adoption as the thing to accelerate.

Why the semiconductor base matters

The strategy is not abstract. It sits on Samsung and SK Hynix, two companies that supply the memory the current AI hardware cycle runs on. That is the mechanism: a state can be growth-first on AI when it already owns a chokepoint in the physical supply chain. Optimism is cheap; a domestic HBM and memory industry is not.

This is the part that does not transfer. A country can copy an AI textbook rollout in a quarter. It cannot copy SK Hynix.

What it signals for RO/EE VCs and founders

The interesting contrast for Eastern Europe is regulatory posture. The EU has chosen the AI Act — risk-tiered, safety-first, compliance-heavy. South Korea's AI Basic Act is the opposite bet: reduce friction, let deployment run, treat AI as industrial policy.

For RO/EE founders this frames a real strategic question. If you build applied AI — eldercare, education, public-sector tooling — the demand signals and the regulatory tailwind are stronger in markets that have chosen the growth-first path. A Romanian eldercare-robotics or edtech-AI startup may find its fastest reference customers in Seoul, not Brussels.

For investors, the read is on public procurement as an anchor customer. South Korea is normalising the state buying AI at scale — textbooks, care robots, national strategy line items. EE governments have the same demographic pressures (ageing populations, care-worker shortages) and far less appetite to spend. The gap between demonstrated demand and public willingness to fund it is where the mispricing sits.

The honest caveat: enthusiasm surveys are soft data. "A majority use AI daily" tells us adoption, not value created or unit economics. Development-over-safety also means South Korea is running an uncontrolled experiment on public-sector AI, and the failures will be public.

Key takeaways

• South Korea's AI Basic Act favours development over safety — a deliberate divergence from the EU AI Act that reframes AI as industrial policy, and a live comparison point for EE regulators.
• The strategy rests on Samsung and SK Hynix, meaning it is defensible in a way pure policy optimism is not — the copyable parts (textbooks, robots) matter less than the chip base.
• State-as-customer is the pattern to watch: EE has the same ageing-population problem but far less public spend, leaving applied-AI founders to look abroad for anchor demand.