The Unjournal · Research prioritization
Paper-specific consideration page · Top AI-related papers, rank 12

Risk Design: AI and Prediction Beyond Screening in Insurance Markets

Limited attention found; quick check
Why this page exists. This paper is on a shortlist of papers we are considering for independent evaluation. No human ratings have been submitted yet, so the score below comes from the AI scoring pass alone. Treat it as one provisional input; the team has made no prioritization decision.
52AI evaluation priority · later Opus assessment
63
Earlier AI score (comparison only)

The combined score uses the later assessment shown alongside. This earlier score is retained for comparison.

None yetHuman aggregate · no ratings submitted
Not availableHuman–AI synthesis needs at least one human rating

Human input

No human ratings have been submitted for this paper yet, so there is no human aggregate or synthesis score. If you know the work, rate or discuss it on the dashboard card; the team reviews feedback before it affects prioritization.

Loading public input and its source labels…

AI assessment: why consider this paper?

The paper models insurance markets where scalable prediction (such as AI) is used to design residual risk and enable prevention, not only to classify fixed risk. It derives a trilemma for low-risk consumers: separate, prevent efficiently, or avoid cross-subsidy, but not all three. It bears on regulation of AI use in insurance. An evaluation would focus on the model's assumptions about AI-treatable risk and empirical grounding.

AI-generated criterion ratings and reasoning

Scores come from AI prioritization; the accompanying explanations include AI-assisted source checks. These are provisional judgments, separate from human ratings and commissioned evaluations.

Decision relevance

5.0/10

The paper addresses a genuinely novel theoretical question: when AI prediction is powerful enough to shape the risks that remain (not just classify pre-existing risk types), does this change welfare analysis of insurance markets? The formal irrelevance result under complete contracting is conceptually important — it establishes a baseline under which AI's expanded role is benign — and the departures from this baseline under incomplete contracting are where policy-relevant market failures emerge. The welfare stakes involve access to insurance and prevention incentives for hundreds of millions of people, primarily in wealthy countries, with a secondary and growing LMIC pathway through insurtech. The incremental contribution over existing work (adverse selection and moral hazard literatures, algorithmic pricing work) lies in the 'risk design vs. risk screening' distinction, which has not been systematically analyzed at this level of generality. The decision-relevance pathway runs through regulatory debates about AI in insurance, which are active in the US, EU, and internationally. The gap between the theory and actionable regulatory guidance is real but not unbridgeable if the paper's incomplete-contracting analysis is well-developed.

  • Paper claim to check Under complete contracting — where prevention is observable, contractible, competitively supplied, and fully priced — it is irrelevant whether AI classifies fixed risk or actively designs residual risk: a Modigliani-Miller-type irrelevance result holds for AI's expanded role in insurance Source abstract

Value of added scrutiny

7.0/10

A trade-press article on AI and risk markets summarises Chan's work, and the paper has an HBS working-paper version. We did two searches and found no independent critique or policy use.

Timing

8.0/10

This is a fresh NBER working paper (w35444 indicates a recent submission in 2025), and the targeted public scrutiny search returned no results — no citations in policy documents, no expert commentary, no PubPeer annotations, no blog discussions. The paper is at maximum timing value for independent evaluation: pre-peer review, no public expert assessment, and in an active policy debate. The absence of search results is uncertainty, not confirmed neglect, but the combination of NBER status and empty scrutiny search suggests the paper has not yet circulated widely beyond the author's network. The AI/insurance regulatory debate is currently live (EU AI Act sectoral implementation, US state insurance commissioner activity, CFPB AI guidance), making a timely expert evaluation particularly valuable if the paper's policy implications are well-developed. Feedback at this stage could still influence framing, policy sections, and the positioning of the incomplete-contracting results before journal submission.

  • Source record Publication-stage and date evidence should be checked in the linked paper record. NBER

Methodological potential

6.0/10

The abstract is truncated and the full paper is paywalled behind NBER, so the complete theoretical model, the characterisation of incomplete-contracting departures, and any calibrated welfare analysis are not directly observable. This is primarily formal economic theory; evaluating it requires experts comfortable with contract theory and IO, who can assess the model assumptions (e.g., market structure, information structures), the generality of the irrelevance theorem, and whether the incomplete-contracting departures are empirically parameterised or purely conceptual. If the paper lacks quantitative estimates or comparative statics tied to real insurance markets, evaluators may find the policy-relevance claims difficult to assess rigorously. The paper's contribution may also be hard to separate from a growing related literature on algorithmic pricing and risk classification (e.g., Einav-Finkelstein, Cohen-Siegelman, Kowalski) — evaluators would need to situate it carefully. The truncated abstract limits confidence in this assessment.

  • Paper claim to check Under complete contracting — where prevention is observable, contractible, competitively supplied, and fully priced — it is irrelevant whether AI classifies fixed risk or actively designs residual risk: a Modigliani-Miller-type irrelevance result holds for AI's expanded role in insurance Source abstract

Prominence

8.0/10

The scoring model estimated prominence from the paper's venue, authors, institutional setting, and visibility. The model did not supply a criterion-specific explanation. Current public-attention status: Limited attention found; quick check.

  • Author-institution working paper HBS hosts a working-paper version. This is the author's institution. Harvard Business School

Likely influence

5.0/10

The scoring model estimated how far the findings could shape later research or decisions, without supplying a criterion-specific explanation. Current public-attention status: Limited attention found; quick check. See public-attention evidence below.

  • Author-institution working paper HBS hosts a working-paper version. This is the author's institution. Harvard Business School

What the paper says

Source abstract · Source-supplied abstract wording; HTML entities and whitespace normalized. Not independently compared with the paper PDF.

I study insurance markets in which scalable prediction, like AI, designs residual risk rather than merely classifies fixed risk. A complete-contracting benchmark shows that if prevention is observable, contractible, competitively supplied, and fully priced, it does not matter whether consumers,

Claims to check

  • Under complete contracting — where prevention is observable, contractible, competitively supplied, and fully priced — it is irrelevant whether AI classifies fixed risk or actively designs residual risk: a Modigliani-Miller-type irrelevance result holds for AI's expanded role in insurance
  • Once contracting is incomplete (prevention unobservable or non-contractible), AI-driven risk design produces qualitatively different welfare and efficiency outcomes than traditional risk screening, disrupting the predictions of classical adverse-selection and moral-hazard theory
  • Scalable AI prediction that shapes risk-prevention behavior (rather than merely sorting pre-existing risk types) generates novel market failures — or potentially welfare improvements — that existing regulatory frameworks designed for actuarial classification are not equipped to address
Methodological or theoretical issues flagged for evaluation

The abstract is truncated and the full paper is paywalled behind NBER, so the complete theoretical model, the characterisation of incomplete-contracting departures, and any calibrated welfare analysis are not directly observable. This is primarily formal economic theory; evaluating it requires experts comfortable with contract theory and IO, who can assess the model assumptions (e.g., market structure, information structures), the generality of the irrelevance theorem, and whether the incomplete-contracting departures are empirically parameterised or purely conceptual. If the paper lacks quantitative estimates or comparative statics tied to real insurance markets, evaluators may find the policy-relevance claims difficult to assess rigorously. The paper's contribution may also be hard to separate from a growing related literature on algorithmic pricing and risk classification (e.g., Einav-Finkelstein, Cohen-Siegelman, Kowalski) — evaluators would need to situate it carefully. The truncated abstract limits confidence in this assessment.

Opus 5.5 re-evaluation (experimental)

Opus raw score: 50/100
Adjusted (+15): 65/100
Opus own action: watchlist
Label from adjusted score: watchlist
Earlier score (gpt-5.5): 76/100
Shift applied to the AI score: -11

Read with care. This is an experimental re-evaluation by a different model (Claude Opus 5.5, 2026-10-01). Across a calibration sample Opus scored about 15 points lower than GPT-5.5, so 15 points are added to compare it with GPT-5.5 scores. The calibration is a small sample (n=44) and is not human-validated, and the two models disagree about the order of papers within the top group. The AI score at the top of this page is the usual evaluation-priority score shifted by the difference between the adjusted Opus score and the earlier holistic score; the synthesis uses that shifted value, and the unshifted score is shown beside it. On the dashboard you can switch the Opus scores off. Read the methods note.

Opus rationale (AI-generated)

This is a recent single-author NBER working paper on how AI prediction interacts with prevention in insurance markets. The topic is clearly in Unjournal's scope (health and insurance economics, social and economic impacts of AI), and as an NBER paper there is no question of field fit. The central argument is that when contracts can change residual risk, competition forces a 'risk-design trilemma', and AI screening can leave society with more avoidable losses, not just different premiums. That speaks to live regulatory choices: CMS risk adjustment in Medicare Advantage and the ACA, state and NAIC rules on algorithmic underwriting, and whether coverage mandates or high-risk subsidies should be justified partly on prevention grounds. It has not been through peer review, and our targeted search found no independent expert discussion, so an evaluation would add real value. Evaluators could check whether the propositions are genuinely new compared with the selection-on-moral-hazard and cream-skimming literatures, how robust they are to the contracting and certification assumptions, and whether any quantification supports the policy claims. Concerns: it looks primarily theoretical, the welfare stakes are mostly in rich-country health insurance, and the transfer pathway to LMIC or other high-priority populations is unclear. For Unjournal this sits in the monitor zone. It is worth commissioning if the full paper includes an empirical calibration or if it starts appearing in policy debates on AI in insurance.

Dashboard details and provenance

Discovery source: NBER
Scoring model: claude-sonnet-4-6 (headless)
Model holistic score: 65.0

Full AI dashboard scoring rationale

This is a solid NBER working paper by Alex Chan that introduces a theoretically novel distinction — AI as risk designer rather than risk screener — with genuine relevance to active regulatory debates about AI in insurance markets. The irrelevance theorem provides a conceptually clean benchmark, and the policy action lies in the incomplete-contracting departures, which map onto the real regulatory questions facing the NAIC, EU, and CFPB. As an unpublished NBER working paper it has received no independent peer review, and the targeted scrutiny search found no prior expert assessments, so evaluation would add real marginal value in checking the model assumptions and the policy relevance of the incomplete-contracting analysis. The case for prioritization now is limited by three factors: the abstract is truncated so the full theoretical and empirical content is uncertain; the welfare pathway is predominantly wealthy-country and indirect; and the paper sits at the theoretical end of a literature that already has active contributors. A 'monitor' recommendation is appropriate — commission an evaluation if the paper gains traction in regulatory or policy circles, or if a closer read of the full paper reveals a stronger calibrated or applied component than the abstract suggests.

AI decision-relevance rationale

The paper addresses a genuinely novel theoretical question: when AI prediction is powerful enough to shape the risks that remain (not just classify pre-existing risk types), does this change welfare analysis of insurance markets? The formal irrelevance result under complete contracting is conceptually important — it establishes a baseline under which AI's expanded role is benign — and the departures from this baseline under incomplete contracting are where policy-relevant market failures emerge. The welfare stakes involve access to insurance and prevention incentives for hundreds of millions of people, primarily in wealthy countries, with a secondary and growing LMIC pathway through insurtech. The incremental contribution over existing work (adverse selection and moral hazard literatures, algorithmic pricing work) lies in the 'risk design vs. risk screening' distinction, which has not been systematically analyzed at this level of generality. The decision-relevance pathway runs through regulatory debates about AI in insurance, which are active in the US, EU, and internationally. The gap between the theory and actionable regulatory guidance is real but not unbridgeable if the paper's incomplete-contracting analysis is well-developed.

AI timing assessment

This is a fresh NBER working paper (w35444 indicates a recent submission in 2025), and the targeted public scrutiny search returned no results — no citations in policy documents, no expert commentary, no PubPeer annotations, no blog discussions. The paper is at maximum timing value for independent evaluation: pre-peer review, no public expert assessment, and in an active policy debate. The absence of search results is uncertainty, not confirmed neglect, but the combination of NBER status and empty scrutiny search suggests the paper has not yet circulated widely beyond the author's network. The AI/insurance regulatory debate is currently live (EU AI Act sectoral implementation, US state insurance commissioner activity, CFPB AI guidance), making a timely expert evaluation particularly valuable if the paper's policy implications are well-developed. Feedback at this stage could still influence framing, policy sections, and the positioning of the incomplete-contracting results before journal submission.

Earlier public-scrutiny search

The targeted scrutiny search (OpenAlex citations, Bing RSS, PubPeer) returned no results for this paper. No independent expert reviews, replication attempts, policy citations, or public commentary were found. The Hypothes.is public-layer annotation check on the NBER URL also returned nothing. This is consistent with a recently released NBER working paper that has not yet circulated widely. Empty search results are uncertainty rather than confirmation of zero scrutiny — the paper may have been discussed at conferences or in working-paper circulation — but the absence of any traceable signal suggests that, at minimum, no substantial public expert assessment has been published. The consequential remaining gap: the paper's irrelevance theorem and its incomplete-contracting extensions have not been independently checked for model assumptions, generality, or policy implications. A commissioned evaluation would be the first rigorous public assessment.

No supporting source links are stored.

Intake, review, and crux connections

Community crux

Can liability insurance mandates deter catastrophic AI risk when developers could not pay for worst-case harms? · 48% match

Insurance-market model analyzes when coverage and prediction incentivize prevention versus moral-hazard or adverse-selection failures.

Public attention and use

AI-assisted review of the linked public sources.

This was a quick check (a few targeted searches). Treat the gaps below as especially uncertain.

What the search did not establish

  • This was a quick check with two targeted searches; treat the gaps as especially uncertain.
  • No EA Forum, LessWrong, regulator or insurer response surfaced.
  • The most decision-relevant next signal is whether insurance regulators engage with the prevention-versus-selection tradeoff.

Quick search checked 2026-10-01. Search scope: Targeted exact-title searches across the open web, institutional and author pages, news and public social-media results, EA Forum/LessWrong, and policy/white-paper contexts. Evidence records distinguish commissioning or report use from independent discussion, media attention, indexing, and post-publication policy use. A search miss is reported as uncertainty, not proof of absence. Entries with check_depth "quick" rest on roughly one to two searches and should be read as especially uncertain; entries checked on 2026-10-01 were added for the shortlist of papers without human ratings.