Skip to content
Apart Research
Apart Research
Contact us
Sprint projectFeb 1, 2026Raleigh, North Carolina/ Ankara, Turkiye/ Accra, Ghana

LINE⁴: Operationalizing the Edge of AI Risk

Adjoa Sarpong, Joseph Awuah, Amo Joshua Selorm · Team Line4

Submitted to The Technical AI Governance Challenge. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: LINE⁴: Operationalizing the Edge of AI Risk

Open in a new tab (opens in new tab)
Recording (opens in new tab)Code (opens in new tab)More on ai-red-line-tracker.streamlit.app (opens in new tab)
Share

LINE⁴ (Line Four) aggregates official safety evaluations across three frameworks to provide transparent, real-time visibility into AI safety assessments across major labs. The dashboard tracks risks across four critical dimensions: CBRN proliferation, cyber offense capabilities, autonomous replication, and deceptive alignment.

Data Source: Real assessments extracted from official lab System Cards

Anthropic: Responsible Scaling Policy (RSP) with Automation Safety Levels (ASL) OpenAI: Preparedness Framework with severity assessments DeepMind: Frontier Safety Framework with Critical Capability Levels (CCL)

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

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

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical AI safety problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.

  1. The team addresses a real problem with a reasonable idea but the work can only be used as a middle man between internal big labs' evaluations and public risk assessments. I can still imagine it being useful however.

  2. its true that the basic vibe of finding the exchange rate between different labs' disclosure vocabulary of capability levels is important and helpful!

    Don't tell me you used streamlit and nextjs in the writeup. That's not a "method", and it spent valuable wordcount that could've been used to elaborate on how you drafted the exchange rates and LLM-as-judge'd over them (sections 3.1 and 3.2 are in my view the most critical, the parts of the paper that actually matter, and you were way too brief about them. I would've happily read more detail about how gemini auditor was prompted, for instance, plus discussion in 3.1 about how you derived buckets and reasons your bucketing might be wrong.

Cite this project

@misc{sarpong2026line4,
  title = {{LINE⁴: Operationalizing the Edge of AI Risk}},
  author = {Adjoa Sarpong and Joseph Awuah and Amo Joshua Selorm},
  year = {2026},
  month = feb,
  note = {Submitted to The Technical AI Governance Challenge, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/line-operationalizing-the-edge-of-ai-risk-jj9e}},
  url = {https://apartresearch.com/sprints/projects/line-operationalizing-the-edge-of-ai-risk-jj9e}
}

More from The Technical AI Governance Challenge

All 47 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026