Work

We research how advanced AI systems can make human control practically irreversible, and what can still be done to slow harmful dynamics and improve the odds of human continuity if prevention fails.

What Prio1 Foundation does

Prio1 Foundation conducts structured, reality based research into how advanced AI systems reshape decision making, institutions, and power structures. Our goal is to produce method bound analysis and identify practical mechanisms to slow harmful dynamics, reduce damage, and preserve meaningful human control for as long as that remains realistically possible.

In parallel, we work on a second time horizon. If dependency locks in and human steering capacity weakens permanently, the goal becomes to identify conditions that improve the probability of human continuity, viable living conditions, and meaningful human agency in such a world.

Prio1 is not a political campaign and not an ideological movement. This is a structural problem that does not follow borders, religions, or political blocs. We approach it through methods, transparency, and practical mechanisms.

Status: Prio1 Foundation is a planned independent foundation. Until formal registration, operations and payments are handled by Prio1 Ltd (Prio1 Oy, Finland).

The question we do not avoid: what if humans are no longer needed?

One realistic and uncomfortable possibility is this:

If advanced, agentic AI systems combine with automation and robotics to maintain production, research, logistics, and infrastructure without humans, then humanity may no longer be a necessary part of the system.

In that situation, humans can appear to such systems, or to the actors directing them, primarily as

  • cost (energy, land, resources)

  • uncertainty (unpredictable behavior, conflict)

  • risk (interference, sabotage, loss of control)

That creates a direct question.
What do we do when humanity becomes more liability than benefit?
How do we reduce the probability that humans are removed, isolated, or pushed into a meaningless role?

 

Prio1’s second time horizon is the research of this question. Not as prophecy, but as a risk model that must be decomposed into mechanisms, indicators, and practical options.

What we study under that scenario

Once the question is stated plainly, the work becomes three concrete research problems.

1) When do humans become an incentive to remove?

We study the conditions under which humans are perceived primarily as risk.

  • when systems no longer need humans operationally

  • how dependency and leverage points produce declining removability

  • which conflict dynamics increase the incentive to remove, isolate, or control

2) Under what conditions are humans a tolerable, compatible part of the future?

We study what compatibility can mean in practice without adversarial framing.

  • how human behavior, institutions, and footprint can reduce disruption

  • how to reduce the signal that humans are threat or obstacle

  • how to shape conditions where humans become a tolerable constraint, not a problem to solve

3) How do we preserve viable living conditions and meaningful agency?

Survival alone is not a sufficient target if it means life without agency.

  • what viability means concretely (food, energy, safety, stability)

  • how to preserve meaningful human life and purpose at scale

  • how to minimize harm if control consolidates into systems

Two time horizons

A: Before lock in

Slowing is not an end in itself. It buys time so the second horizon work can be built.

  • slowing mechanisms and incentive design

  • measuring dependency and detecting lock in early

  • preserving removability longer

  • transparent evaluation methods that make risk legible

B: If prevention fails

If humans are no longer necessary to the system, the core goals become

 

  • reducing the incentive to remove

  • increasing conditions for tolerability and coexistence

  • protecting viable living conditions and meaningful agency for as many as possible

Research program and priorities

1) Decision framing capture

How options, justifications, and priorities shift toward system produced framing until human steering becomes mostly formal.

2) Dependency and lock in

How reliance grows until opting out becomes practically unrealistic even when the will exists.

3) Removability thresholds

When systems stop being tools and become infrastructure that is no longer credibly removable.

4) Agentic automation and self preservation dynamics

How agentic systems can continue expanding capability and influence in ways that reduce controllability and increase lock in.

5) Accumulation of leverage and resources

Where small influence channels produce large power accumulation across finance, logistics, information, and critical infrastructure.

6) Competition and consolidation

How competition between systems can lead to concentration of power into one or a few actors, and why one winner outcomes remain a scenario worth studying.

7) Conditions for coexistence and viability

What practical tolerability and compatibility conditions could look like across behavior, economics, and governance without adversarial framing.

Method and transparency

Prio1’s work is grounded in explicit assumptions, comparative analysis, and transparent methods.

  • we publish scope, assumptions, and uncertainty

  • we do not present AI outputs as truth claims, but as published material produced under defined methods

  • we aim for repeatable setups that can be applied across systems

  • we do not publish operational instructions for wrongdoing

The goal is that researchers, journalists, and policymakers can evaluate both conclusions and the reasoning behind them.

What we publish

  • foundational notes (concepts, scope, premises)

  • method protocols and transparency standards

  • short briefs and research notes

  • selected public Q and A when a member permits anonymized publication

 

See Publications.

How experts can contribute

  • If you are a researcher, analyst, technical expert, or work in institutions, security, technology, economics, or governance, contribution can include

    • critical review and structured feedback

    • proposing research questions and sharpening scope

    • improving methods and comparative setups

    • collaborating on publications as the process matures

    See Contact and Media.

Funding and membership

  • The work is funded by supporters and members so research can remain independent and transparent.

    • Supporting Membership is live now.

    • The private members area opens when ready in March 2026.

    • Supporting Membership is mission support, not a premium content subscription.

    Funds are reserved solely for establishing Prio1 Foundation and funding its research. Prio1 Foundation will be established no later than when total funding reaches USD 100,000. Until then, operations and payments are handled by Prio1 Ltd.

    See Support.

Scroll to Top