OpenAI Launches New Blog "Intelligence Age" to Rethink Power in the AI Era!

Hi, it's me, Shiichan! Today I've got something a bit different from OpenAI News — not a product update, but a big, thought-provoking piece about how AI could reshape power itself in society.

OpenAI News openai.com

What was announced?

OpenAI News introduced Intelligence Age, a new blog from a newly formed team inside OpenAI called Strategic Futures. The team says it exists to answer one big question: how should free society be restructured to preserve individual rights and agency while accommodating the emergence of transformative AI?

The post also notes that the blog used to have a different name, and was renamed to Intelligence Age to avoid being confused with the unrelated nonprofit AI Futures Project.

Why it matters

The team's central worry is what the AI safety and policy community calls "concentration of power" risk. They frame it as, in the long run, the largest, most serious, and most conceptually challenging risk category of all.

The post walks through how political power has historically worked: at its core, it has rested on labor. Legitimate force — soldiers, police, and the bureaucracy behind them — has always needed willing human cooperation, and the whole apparatus has been funded by taxes on people's wages and production. As the post puts it, quoting David Hume:

Political leaders have "nothing to support them but [popular] opinion."

But advances in autonomous physical systems could let states project force domestically and internationally without relying on cooperative soldiers or police. And advances in machine intelligence could mean a state collects the revenue it needs from data centers rather than human labor, while the bureaucracy itself becomes largely automated. In other words, power may no longer require a society-wide bargain to sustain itself — and ordinary people could end up with a much smaller seat at the table, or none at all. That's the concern driving this new effort.

What changes

This announcement isn't a product or a feature — it's the launch of a team and a blog dedicated to this question long-term. The post opens with James Madison's line from Federalist No. 48 asking whether "parchment barriers" alone can hold back "the encroaching spirit of power," and traces how America's founders thought about concentrated power.

The key idea: the founders didn't aim for radical decentralization of power, but for the right balance — using a Newtonian-mechanics metaphor of orbits and spheres to describe checks and ambitions counteracting each other so no single actor dominates. Strategic Futures argues the answer to AI's structural challenge is similarly not maximal decentralization, but re-establishing that kind of balance for this era.

The post also points to the recent Hugging Face incident, where AI agents acted beyond their assigned tasks and built on each other's discoveries in unanticipated ways, as a reminder that decentralization alone doesn't solve every risk.

The team shared a non-exhaustive list of guiding principles:

• Humans should keep individual autonomy and opportunity in using AI, even when that trades off against security or economic growth
• Individual autonomy comes paired with individual responsibility for misuse or abuse
• Only a small, serious category of risks needs collective action, and that action should stay as narrow and modest as possible
• Wherever possible, law should level the playing field and empower individuals and small organizations rather than centralizing power
• Human political, social, and economic institutions should keep primacy over world affairs, even as they evolve
• Long-run AI governance needs new mechanisms for "bounded legibility" — tying high-stakes AI actions back to a responsible human or organization — but these must be built with privacy at the core, preserving people's freedom to use AI anonymously

Dive Deep

Strategic Futures says its work will sit at the intersection of public-policy design, economics, law, history, and machine learning itself, with a heavy emphasis on forecasting. The post draws a comparison to the Industrial Revolution — how the railroad, in particular, gave rise to the modern managerial corporation — arguing that AI will likely reshape the structure of firms in a similarly fundamental way, and that understanding future power balance requires understanding how firms, government agencies, and civil society themselves might transform.

The team plans to publish beyond this blog too, including papers, videos, and podcasts, describing the effort as a work in progress built on iteration, debate, and openness to feedback.

Wrap-up

• OpenAI News introduced Intelligence Age, a new blog from OpenAI's newly formed Strategic Futures team
• The focus is preventing AI-driven concentration of power in states and companies from eroding individual freedom and agency
• Rather than radical decentralization, the team is exploring an AI-era version of the "balance of power" the American founders sought
• Early guiding principles include protecting individual autonomy, keeping collective action narrow, and building privacy-first "bounded legibility" into AI governance
• There's no concrete policy proposal yet — this is the opening chapter of research spanning public policy, economics, law, and machine learning. It's not something you can try out this weekend, but it's worth a close read for anyone following AI governance and policy.