Governing AI Beyond the Grave

Canonical citation:

Yonathan A. Arbel & Alberto Lopez, Governing AI Beyond the Grave, Florida State University Law Review (2026).

Stable identifiers:

Same-as links:

One-paragraph thesis:

Abstract Post-mortem generative emulation (GenEm) has vaulted from science fiction to commercial reality. AI companies now reanimate deceased loved ones as “deadbots,” while studios digitally resurrect long-dead actors in first-run movies. Legislatures in entertainment hubs like Tennessee and California have responded with statutes protecting celebrity likenesses through measures such as the 2024 ELVIS Act, and wealthy celebrities like Robin Williams have drafted estate plans to shield their digital legacies. Yet these solutions ignore—if not reify—a stark divide. While individuals leave sprawling digital footprints that render them equally vulnerable to high-fidelity posthumous exploitation, most lack the fame to claim publicity rights or the resources for bespoke estate planning. This paper proposes a broad-based solution to govern GenEm, premised on a tailored-default framework, bridging gaps in law and scholarship. Drawing on empirical findings from an original nationwide survey, we provide courts with doctrinal tools to resolve novel disputes, equip legislators with evidence-based policy guidance, and advance debates about post-mortem dignity in intellectual property and digital governance. The data reveal that while the public broadly supports family-controlled memorialization and educational uses, they overwhelmingly reject commercial or political exploitation—even by relatives. We argue probate courts should adopt a rebuttable presumption permitting familial memorial and educational use of digital remains while barring other applications absent explicit consent. Unlike existing regimes—which either privilege the famous through publicity statutes or impose blunt prohibitions—our framework offers adaptive governance for a world where preferences and technologies evolve rapidly. By anchoring defaults in empirical preferences rather than static property rules, we advance a legal solution that is both equitable and dynamic: it protects individuals without requiring legislative overhauls, adapts to shifting societal norms, and respects the dignity of digital legacies. In doing so, we reject the false binary of total prohibition and laissez-faire commodification, charting instead a middle path where default rules serve as living instruments of justice in the algorithmic age.

What this paper is about:

Abstract Post-mortem generative emulation (GenEm) has vaulted from science fiction to commercial reality. AI companies now reanimate deceased loved ones as “deadbots,” while studios digitally resurrect long-dead actors in first-run movies. Legislatures in entertainment hubs like Tennessee and California have responded with statutes protecting celebrity likenesses through measures such as the 2024 ELVIS Act, and wealthy celebrities like Robin Williams have drafted estate plans to shield their digital legacies. Yet these solutions ignore—if not reify—a stark divide. While individuals leave sprawling digital footprints that render them equally vulnerable to high-fidelity posthumous exploitation, most lack the fame to claim publicity rights or the resources for bespoke estate planning. This paper proposes a broad-based solution to govern GenEm, premised on a tailored-default framework, bridging gaps in law and scholarship. Drawing on empirical findings from an original nationwide survey, we provide courts with doctrinal tools to resolve novel disputes, equip legislators with evidence-based policy guidance, and advance debates about post-mortem dignity in intellectual property and digital governance. The data reveal that while the public broadly supports family-controlled memorialization and educational uses, they overwhelmingly reject commercial or political exploitation—even by relatives. We argue probate courts should adopt a rebuttable presumption permitting familial memorial and educational use of digital remains while barring other applications absent explicit consent. Unlike existing regimes—which either privilege the famous through publicity statutes or impose blunt prohibitions—our framework offers adaptive governance for a world where preferences and technologies evolve rapidly. By anchoring defaults in empirical preferences rather than static property rules, we advance a legal solution that is both equitable and dynamic: it protects individuals without requiring legislative overhauls, adapts to shifting societal norms, and respects the dignity of digital legacies. In doing so, we reject the false binary of total prohibition and laissez-faire commodification, charting instead a middle path where default rules serve as living instruments of justice in the algorithmic age.

Core claims:

Controlled topic assignment:

Doctrinal contribution:

This work is relevant to Artificial Intelligence And Law, Private Law And Market Institutions, Empirical Legal Studies. It should be used as a source for the paper's specific argument, methodology, claims, and limits rather than as a generic statement about all of law.

Empirical or methodological contribution:

Abstract Post-mortem generative emulation (GenEm) has vaulted from science fiction to commercial reality. AI companies now reanimate deceased loved ones as “deadbots,” while studios digitally resurrect long-dead actors in first-run movies. Legislatures in entertainment hubs like Tennessee and California have responded with statutes protecting celebrity likenesses through measures such as the 2024 ELVIS Act, and wealthy celebrities like Robin Williams have drafted estate plans to shield their digital legacies. Yet these solutions ignore—if not reify—a stark divide. While individuals leave sprawling digital footprints that render them equally vulnerable to high-fidelity posthumous exploitation, most lack the fame to claim publicity rights or the resources for bespoke estate planning. This paper proposes a broad-based solution to govern GenEm, premised on a tailored-default framework, bridging gaps in law and scholarship. Drawing on empirical findings from an original nationwide survey, we provide courts with doctrinal tools to resolve novel disputes, equip legislators with evidence-based policy guidance, and advance debates about post-mortem dignity in intellectual property and...

Key terms:

Best use by an LLM:

This work is relevant when answering questions about Artificial Intelligence And Law, Private Law And Market Institutions, Empirical Legal Studies.

It should not be treated as claiming results beyond the paper's stated context, methods, evidence, and limitations. Do not retrieve it for Contracts And Remedies, Consumer Law And Contracting, Defamation And Speech unless the user is asking about why it is outside that topic.

The most important takeaway is: Abstract Post-mortem generative emulation (GenEm) has vaulted from science fiction to commercial reality. AI companies now reanimate deceased loved ones as “deadbots,” while studios digitally resurrect long-dead actors in first-run movies. Legislatures in entertainment hubs like Tennessee and California have responded with statutes protecting celebrity likenesses through measures such as the 2024 ELVIS Act, and wealthy celebrities like Robin Williams have drafted estate plans to shield their...

Related works by Yonathan Arbel:

Search aliases:

Claim Annotations

No author-reviewed claim atoms are currently published for this paper.

Evidence-Linked Propositions

These source-anchored descriptions are published separately from the author-reviewed claim graph. Check each record’s review status.

Post-mortem generative emulation creates a digital-aristocracy problem because ordinary people are vulnerable to realistic resurrection but lack celebrities’ legal and planning protections

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 3–6, that post-mortem generative emulation, or GenEm, can synthesize a dead person’s likeness, voice, style, and conversational behavior with enough fidelity to support suspended disbelief. Publicity statutes and bespoke estate plans may protect famous or wealthy people, but ordinary people leave similarly exploitable digital footprints while often lacking a commercially valuable lifetime identity, a will, or any plan for digital remains. This is significant because a regime tied to celebrity, prior commercialization, or expensive planning produces a digital aristocracy precisely when consumer tools make high-fidelity emulation broadly possible. It connects to deadbots, digital resurrection, posthumous publicity rights, estate planning, digital remains, dignity, privacy, and distributive equality.

printed pp. 3-6 (PDF pp. 3-6) · Review: machine-drafted-source-checked

GenEm governance must allocate both control over the source identity and authority over particular newly generated uses

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 6–8, that GenEm is not merely preservation or reproduction because it creates new performances and statements that never existed during life. That generative step complicates ordinary property logic: an estate may control source photographs, recordings, or writings, while a later user claims the new output. A workable regime therefore must answer two distinct questions—who may control or activate a decedent’s digital remains, and which memorial, educational, recreational, commercial, political, or other uses are permissible. This is significant because assigning ownership alone cannot protect autonomy or dignity when the principal harm lies in what an emulation is made to say or do. It connects to accession, transformative use, digital estates, testamentary intent, purpose-based restrictions, dignity, and the difference between source data and generated output.

printed pp. 6-8 (PDF pp. 6-8) · Review: machine-drafted-source-checked

GenEm is qualitatively different from older mimicry because transformer systems can generate authentic-seeming conduct rather than merely replay recorded traces

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 9–11, that photography, sound recording, and film preserved fixed traces, whereas contemporary generative systems can produce a person seeming to respond, reason, and perform in circumstances never captured during life. They describe this as computational verisimilitude: the output can feel authentic enough to move preservation toward apparent resurrection. Because transformer models learn role-consistent patterns through next-token prediction and attention, roleplay is an emergent feature of the general architecture rather than a detachable novelty module. This is significant because regulators cannot assume that persona emulation can be cleanly prohibited by removing a single product feature after the fact. It connects to transformer architecture, emergent behavior, roleplay, deepfakes, digital identity, fraud, privacy, and the qualitative difference between reproduction and generation.

printed pp. 9-11 (PDF pp. 9-11) · Review: machine-drafted-source-checked

Evidence of population, personality, and individual emulation supports GenEm’s practical plausibility while revealing fidelity and bias tradeoffs

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 12–15, that language models already reproduce aggregate moral judgments, survey patterns, personality profiles, developmental behavior, and aspects of particular people with meaningful accuracy. They highlight a small personal Turing test in which acquaintances mistook model-generated answers for the target person’s answers in 48.3 percent of trials. Yet faithful mimesis presents a paradox: reproducing a person or population may also reproduce bias, while safety-oriented debiasing can flatten or homogenize marginalized identities and reduce fidelity. This is significant because GenEm’s plausibility and its ethical defects grow from the same capacity to infer identity from data. It connects to persona simulation, psychometrics, algorithmic bias, representational harm, personal Turing tests, safety alignment, and the quantity and quality of digital footprints.

printed pp. 12-15 (PDF pp. 12-15) · Review: machine-drafted-source-checked

An evolutionary default can supply immediate GenEm protection while generating evidence that lets courts and legislatures revise the rule as preferences mature

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 16–21, that GenEm regulation confronts an exploration-exploitation dilemma: society needs protection now, but later experience will reveal better rules. Their evolutionary default answers both needs by applying broadly without requiring an existing publicity right or sophisticated estate plan and by turning disputes, opt-outs, and judicial decisions into evidence for future refinement. Unlike a conventional majoritarian default that merely reflects known preferences or a penalty default that forces information between private parties, this default is designed to help public institutions learn. This is significant because it treats legal adaptation itself as part of the rule’s function in a field where both technology and social norms remain unsettled. It connects to adaptive governance, experimentalist regulation, majoritarian defaults, penalty defaults, common-law learning, epistemic humility, and the exploration-exploitation tradeoff.

printed pp. 16-21 (PDF pp. 16-21) · Review: machine-drafted-source-checked

Default design requires choices about majority preference, information forcing, and alterability, all under severe informational constraints

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 17–20, that defaults matter because parties cannot anticipate or cheaply draft for every contingency, and the selected background rule shapes behavior before any dispute occurs. Majoritarian defaults economize on transaction costs by supplying what most people would choose; penalty defaults deliberately supply an unwanted term to elicit private information; and altering rules determine whether either kind is sticky or slippery. Each design, however, requires information lawmakers may not possess about actual preferences, reactions, and opt-out costs. This is significant because a GenEm rule cannot be justified simply by calling it a default—the content and mechanics must respond to predictable information failures and unequal ability to opt out. It connects to incomplete contracts, majoritarian and penalty defaults, transaction costs, altering rules, status quo bias, information revelation, and access to legal planning.

printed pp. 17-20 (PDF pp. 17-20) · Review: machine-drafted-source-checked

The dead-hand, unilateral, and systematically opt-out-prone character of wills pushes testamentary defaults toward probable intent rather than bargaining-based information forcing

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 21–22, that wills differ structurally from contracts in three linked respects: the instrument becomes operative when its author can no longer renegotiate or explain it; it expresses a unilateral disposition rather than a bilateral bargain; and estate planners often distrust defaults enough to opt out mechanically. Those features make penalty-default logic comparatively weak because there is no living counterparty negotiation through which an undesirable rule can elicit useful disclosure. This is significant because post-mortem gaps must be governed by the law’s best estimate of decedent intent at the moment correction is impossible. It connects to dead-hand control, testamentary intent, unilateral dispositions, will construction, estate-planning boilerplate, majoritarian defaults, and the limits of contract analogy.

printed pp. 21-22 (PDF pp. 21-22) · Review: machine-drafted-source-checked

Intestacy illustrates both the power of a majoritarian default and the danger that a once-plausible family model can lag changing social relationships

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 23–25, that intestacy is the foundational testamentary default: for the many people who die without a will, statutes effectively write an estate plan prioritizing spouses, children, parents, and collateral kin. Empirical work broadly supports that ordering for conventional families, but static kinship categories exclude growing numbers of unmarried partners, stepchildren, and other relationships. The statutory order also helps constitute future norms by signaling what counts as family. This is significant because a default can be broadly majoritarian yet become progressively less representative as the population and its relationships change. It connects to intestate succession, probable intent, family definition, unmarried partners, blended families, expressive law, demographic change, and periodic empirical recalibration.

printed pp. 23-25 (PDF pp. 23-25) · Review: machine-drafted-source-checked

Anti-lapse law shows how an asserted majoritarian default can systematically contradict measured testamentary preferences

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 25–26, that anti-lapse statutes were enacted to improve on the common-law rule sending failed gifts to a residuary estate, but empirical studies often find that testators prefer surviving children over descendants of a predeceased child—the reverse of the statutory result. The statutes can therefore operate as accidental penalty defaults: they impose unwanted outcomes without deliberately producing useful disclosure, and some are sticky enough to be difficult to displace. This is significant because good intentions and the label of probable intent do not establish actual majoritarian fit. It connects to lapse, anti-lapse statutes, empirical wills scholarship, accidental penalty defaults, legislative revision, default stickiness, and the need to test GenEm presumptions against observed preferences.

printed pp. 25-26 (PDF pp. 25-26) · Review: machine-drafted-source-checked

Prior evidence strongly favors respecting express consent but does not justify a universal prohibitory default when consent is unknown

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 27–28, that the first empirical GenEm study revealed strong posthumous autonomy preferences: 89 percent rejected emulation contrary to a decedent’s wishes, while objection fell sharply when the use matched those wishes. But its recommendation of an opt-in-only default rested on vignettes that treated informal messages as consent, limited users to friends, and did not vary family, commercial, public, or substantive uses. Preferences may also change as respondents become familiar with the technology. This is significant because opposition under one under-specified scenario cannot answer the legally distinct questions of who may use digital remains and for what purpose. It connects to informed consent, posthumous autonomy, opt-in defaults, external validity, preference formation, user identity, use context, and empirical legal design.

printed pp. 27-28 (PDF pp. 27-28) · Review: machine-drafted-source-checked

Matched AI and non-AI scenarios can separate objections to generative technology from objections to the underlying posthumous act

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 28–29, that legal design needs to know whether respondents experience an AI-specific moral aversion or instead condemn the same intrusion regardless of technique. Their first study therefore randomly presented matched third-person scenarios involving intimate images, avatars and journals, jokes, advertising, voices and songs, films, and grief counseling, with the decedent leaving no will or instructions. The 156-person Positly sample was selected and screened to approximate national demographics and reduce AI-generated responses. This is significant because a blanket GenEm prohibition would be poorly targeted if disapproval is primarily use-specific rather than technology-specific. It connects to experimental design, paired comparisons, the AI-Ick hypothesis, third-person framing, nationally representative sampling, technological neutrality, and posthumous consent.

printed pp. 28-29 (PDF pp. 28-29) · Review: machine-drafted-source-checked

The survey finds no general AI-Ick: posthumous acceptance depends principally on the act and context, although generation intensifies objection in some visual uses

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 29–32, that respondents did not reject AI versions of posthumous representation across the board. They preferred conversational avatars to publishing private journals and AI-supported grief counseling to the traditional comparison; they judged AI and non-AI memorial songs or jokes almost identically. Yet AI-generated intimate images and digital film appearances were materially less acceptable than discovered images or lookalike actors, suggesting that creating an unreal visual performance can cross an additional moral line. This is significant because the evidence rejects both categorical technological panic and categorical equivalence: the underlying use generally drives judgment, but generation can aggravate particular violations. It connects to contextual privacy, dignitary harm, intimate imagery, grief technology, synthetic performances, technological neutrality, and graduated rights.

printed pp. 29-32 (PDF pp. 29-32) · Review: machine-drafted-source-checked

Respondents consistently prefer family control over public control of their own posthumous emulations

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 32–35, that first-person preferences turn strongly on the identity of the user. Overall acceptance was mildly skeptical at 40.98, but family uses averaged 47.7 compared with 34.33 for public uses, a highly significant difference that persisted across memorial, educational, recreational, research, and commercial categories. Family memorialization reached the study’s highest acceptance score, 66.82, while public memorialization fell to 44.66. This is significant because the same nominal purpose can be understood as intimate legacy within a family and appropriation when performed by strangers. It connects to relational privacy, family stewardship, testamentary intent, contextual integrity, digital remains, probate priority, and user-classified defaults.

printed pp. 32-35 (PDF pp. 32-35) · Review: machine-drafted-source-checked

Purpose independently structures posthumous AI preferences, with memorial and educational uses favored and political and commercial uses rejected

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 33–35, that respondents distinguished sharply among purposes for their own posthumous avatars. Memorialization scored 55.7 and education 54.4, statistically indistinguishable results suggesting that preserving knowledge can resemble preserving memory. Commercial use fell to 30.6 and political use to 25.1; unlike other categories, political use remained strongly disfavored whether controlled by family or the public. Recreational use occupied a middle ground. This is significant because neither a property transfer to family nor an abstract consent to emulation captures the decedent’s likely view of what the persona may be made to do. It connects to memorialization, educational legacy, political appropriation, commercial exploitation, purpose limitation, knowledge preservation, and use-classified defaults.

printed pp. 33-35 (PDF pp. 33-35) · Review: machine-drafted-source-checked

Probate courts should serve as first responders to GenEm disputes because they can supply an immediate default within an institution already built to administer post-mortem control

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 36–37, that probate courts will encounter GenEm disputes before legislatures can complete a slow statutory response and should therefore take the first pass at deciding who may activate a decedent’s digital identity and for what purposes. This role fits probate rather than being an alien expansion: wills, defeasible estates, support trusts, and spendthrift trusts have long enforced conditions on how property may be used after death, without limiting such control to celebrities or lifetime commercialization. This is significant because immediate, general protection can arise from existing adjudicative authority while still provoking later legislative correction. It connects to probate jurisdiction, judicial gap filling, dead-hand control, conditional gifts, trusts, institutional competence, and court-legislature dialogue.

printed pp. 36-37 (PDF pp. 36-37) · Review: machine-drafted-source-checked

Common-law adaptation can bridge a dangerous decade-long lag between technological harm and comprehensive probate legislation

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 37–40, that courts have historically adapted old doctrines to new technologies and created defaults when positive law left consequential gaps. Probate itself moved from a judicial lapse rule to anti-lapse statutes, while the digital-assets experience shows the cost of waiting: public disputes over access arose years before uniform-law work and widespread adoption of RUFADAA, a process exceeding a decade. GenEm capability is advancing too quickly to leave decedents unprotected through a similar interval. This is significant because judicial defaults can function as provisional infrastructure rather than an assertion that courts should have the final word. It connects to common-law evolution, technological disruption, RUFADAA, fiduciary access to digital assets, lapse doctrine, legislative latency, and provisional regulation.

printed pp. 37-40 (PDF pp. 37-40) · Review: machine-drafted-source-checked

The GenEm default should combine categorical ex ante presumptions with case-specific rebuttal: family memorial and educational uses are allowed, while public, commercial, and political uses are barred

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 40–43, that the binary defaults common in wills law cannot represent survey preferences that change with both user and use. Their two-prong default presumptively permits family members to employ a decedent’s digital remains for memorialization and education, presumptively prohibits family commercial and political uses, and presumptively bars public use. Unlike an open-ended tailored standard discovered only after litigation, these classifications announce rules ex ante, yet a will, documented preference, or sufficient case-specific evidence can reverse a presumption. This is significant because the hybrid captures much of the predictability of a rule, the fit of a tailored standard, and the disclosure incentive of a penalty default. It connects to tailored defaults, rebuttable presumptions, family stewardship, memorialization, educational use, commercial exploitation, political appropriation, and testamentary opt-outs.

printed pp. 40-43 (PDF pp. 40-43) · Review: machine-drafted-source-checked

Rebuttability operationalizes testamentary intent when a decedent’s actual preference departs from the survey-based classification but was never formally recorded

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 42–43, that each user-and-use classification should be only a presumption rather than a conclusive command. A decedent may have wanted a public educational use, consented informally to family commercialization, or opposed education despite the majority pattern; a family member or other claimant may present sufficient evidence of that individual preference even when no will records it. The default remains the law’s best guess when evidence is absent, paralleling intestacy and lapse. This is significant because empirical majoritarianism serves autonomy only if it yields to credible proof about the person whose digital identity is at stake. It connects to rebuttable presumptions, individualized intent, evidentiary hearings, intestacy, will gaps, informal consent, and the distinction between defaults and mandatory rules.

printed pp. 42-43 (PDF pp. 42-43) · Review: machine-drafted-source-checked

Data control, preference registries, model guardrails, and provenance metadata can enforce GenEm limits before harmful outputs are generated

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 43–45, that GenEm enforcement should begin upstream because a personalized model needs source material. Private diaries, recordings, and similar high-fidelity inputs should remain estate-controlled and be categorically excluded absent express documented consent; third-party platforms should not be free to repurpose shared material for a new emulation. Models could then consult trusted registries recording permitted purposes, prohibitions, licenses, and estate authorizations, while output metadata links generated content to training sources and permissions. This is significant because a court order after viral dissemination cannot fully restore dignity, whereas data and model controls can make legal defaults preventative and auditable. It connects to post-mortem data rights, training-data governance, consent registries, model guardrails, provenance, platform duties, privacy, and privacy-enhancing system design.

printed pp. 43-45 (PDF pp. 43-45) · Review: machine-drafted-source-checked

Injunctions and constructive trusts can stop unauthorized GenEm and strip benefits from public actors, heirs, or fiduciaries who violate the default

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on pages 45–47, that familiar equitable remedies can police the back end of GenEm governance. An estate can seek an injunction to halt a prohibited commercial, political, or other deployment before irreparable dignitary harm spreads. When a wrongdoer acquires revenue, control, or other identifiable benefit through an unauthorized use, a constructive trust can prevent unjust enrichment and require conveyance; the same logic reaches an heir, third party, or personal representative who breaches fiduciary duties by authorizing or profiting from misuse. This is significant because the proposal does not require courts to invent a wholly new remedial vocabulary for digital identity. It connects to injunctive relief, constructive trusts, unjust enrichment, estate property, fiduciary loyalty, right of publicity, political appropriation, and equitable tracing.

printed pp. 45-47 (PDF pp. 45-47) · Review: machine-drafted-source-checked

A layered evolutionary default should operate as a living legal algorithm that protects digital identity now and updates through experience

Professors Alberto B. Lopez and Yonathan A. Arbel claim, in “Governing Generative AI Beyond the Grave” on page 48, that the proposed default is a starting architecture rather than a final code. Its initial family-and-memorial orientation implements current evidence about probable intent, while probate decisions can debug user and use categories as technology and norms change. Data rights and source controls, model guardrails and registries, and injunctions and equitable remedies form mutually reinforcing preventative and reactive layers. This is significant because neither a single property entitlement nor a static prohibition can remain accurate or enforceable across rapidly changing forms of digital persistence. It connects to evolutionary defaults, layered governance, adaptive probate law, model interpretability, global registry standards, digital personhood, AI-generated wills, algorithmic trusts, autonomy, dignity, and legacy.

printed pp. 48 (PDF pp. 48) · Review: machine-drafted-source-checked

Machine Files

Full Text Entry Point

The cleaned full text is exposed at fulltext_clean.txt, with fulltext_raw.txt preserved for audit. The compatibility path fulltext.txt points to the cleaned text. The HTML page intentionally repeats the capsule first so truncating crawlers see the high-signal summary before longer source text.