Skip to content
Apart Research
Apart Research
Contact us
Sprint projectSep 14, 2025

Foundation

Eduard Kapelko · Team Foundation

Submitted to CBRN AI Risks Research Sprint. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Foundation

Open in a new tab (opens in new tab)
Share

This document presents the conceptual framework for establishing an organization named the "Foundation," whose primary goal is to mitigate the existential and systemic risks associated with advanced artificial intelligence. The text outlines a structure designed to ensure the safe and collaborative international development of Artificial General Intelligence (AGI) and Artificial Superintelligence (ASI), thereby preventing threats such as the realization of the "Vulnerable World Hypothesis," the gradual disempowerment of humanity, and a destabilizing AI arms race. To achieve these goals, a governance model is proposed based on the principles of a narrow mandate, radical financial transparency, rotation of power, and the use of liquid democracy tools in a DAO format. A key security element is a multi-layered internal access control system, consisting of independent teams, which ensures a comprehensive audit of the models and the audit process itself. The document also describes an economic model that transforms the competitive race into a collaborative effort by pooling resources and creating a network effect that incentivizes participants to join the Foundation.

Reviews

Judging this Sprint?

Review “Foundation”

Your public critique appears on the project page without your name. Reviews of strong projects are shared with funders, as you choose below.

Not shown on the project page.

Shown on the project page, without your name.

Only the Apart team reads this, and funders if you agree below.

Decline to score

Only when the quality of the writing makes the work impossible to follow. Say why in your private critique; the project page shows only that a judge declined, not why.

Does the project meaningfully address the theme of CBRN risks and AI misuse? Does the project clearly connect to a CBRN-related threat (chemical, biological, radiological, nuclear)? Does it build on or challenge existing literature or prior work in the space? Does it offer a novel framing, tool, or evaluation that could inform future research or policy? Is it grounded in real-world relevance rather than hypothetical misuse?

Does this project advance the field of AI safety in a clear and valuable way? Does it help identify or mitigate risks specific to powerful AI systems? How well does the project engage with the unique challenges of AI alignment, misuse, or system evaluation? Could the output contribute to scalable safety mechanisms or evaluation pipelines? Does it generalize to broader dual-use or frontier AI concerns?

Is the project thoughtful, usable, and technically sound? Is the technical implementation solid and well-scoped for a research sprint? Is the code or method clearly documented and reproducible? Does the project demonstrate sound methodology, clear assumptions, and an honest treatment of limitations? Would the tool or prototype be useful for future research, governance, or evaluation work?

Sharing your name *

When we share a strong project with funders, your name, scores and public critique go with it, confidentially; this choice is only about whether your name is public. If you think this project deserves funding, please also endorse it: that carries the most weight.

Share my private critique with funders *

  1. Key strengths of the approach

    1. Well-structured and well-fleshed out overall(although a fair bit longer than the 5 page limit!)
    2. Strong thinking on the problems and limitations of Red Team and Purple Team models.
    3. Clear effort to design transparent structures, set proper objectives, and define roles for each team.
    4. Good attention to structural control and incentivization mechanisms, which shows thoughtful design.

    Specific areas for improvement

    1. The project does not directly engage with CBRN risks or explain how it addresses a specific CBRN threat.It also risks being perceived as “just another body to govern AI,” echoing the mythical idea of a single international AI body.
    2. Misses an upfront framing of how the DAO could specifically mitigate CBRN risks and how this adds value compared to existing paradigms.

    Suggestions for how to develop this into a stronger project

    Reframe the DAO explicitly around CBRN functions, for example, highlight that one of its many functions could be to mitigate CBRN risks and provide specificity: what does DAO-based governance look like when facing a biothreat, radiological incident, or chemical accident?

    Potential next steps or future directions

    1. Develop an example case e.g., DAO’s role during an accidental pathogen release, radiological material smuggling).
    2. Explore integration with existing international CBRN regimes (e.g., OPCW, IAEA) rather than aiming to reinvent a broad global AI body which is very tricky. If looking to do a bottom-up approach then try to show how the incentives for the AI companies would be to back this or take part in it(an example could be maybe something like the Frontier Model Forum).
    Read full reviewShow less

Cite this project

@misc{kapelko2025foundation,
  title = {{Foundation}},
  author = {Eduard Kapelko},
  year = {2025},
  month = sep,
  note = {Submitted to CBRN AI Risks Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/foundation-4zzi}},
  url = {https://apartresearch.com/sprints/projects/foundation-4zzi}
}
Browse all projects

A Sprint is running now

AI Collusion Research Sprint · Oct 23 - 25, 2026