Step 4: Full View

Beyond the Board's finding that Engineer A must recommend minimizing harm to the least number of persons, Engineer A bears an additional obligation to explicitly disclose to the automobile manufacturer that this recommendation is grounded in a utilitarian ethical framework rather than in any established regulatory or industry standard. Because no applicable national or industry standards governing autonomous vehicle harm-allocation decision logic currently exist, Engineer A cannot represent the harm-minimization recommendation as a technically mandated or universally accepted engineering norm. Presenting it as such would violate the completeness and non-selectivity obligation that governs Engineer A's advisory role. Engineer A must therefore clearly communicate to the automobile manufacturer that the recommendation reflects a specific moral philosophy - one that reasonable engineers and ethicists might contest - so that the manufacturer can make a genuinely informed deployment decision. This disclosure obligation is heightened, not relieved, by the regulatory standards vacuum, because the absence of external standards places the full burden of ethical transparency on Engineer A as the professional advisor.

The Board's conclusion that Engineer A must recommend minimizing harm to the least number of persons implicitly adopts a utilitarian ethical framework - specifically, an aggregate harm-minimization calculus - without acknowledging that this represents one among several defensible moral philosophies rather than a universally accepted engineering standard. A deontological framework, for instance, might prohibit the vehicle from actively redirecting harm toward any third party regardless of aggregate outcome, treating each person's life as inviolable rather than as a unit in a welfare sum. Because Engineer A is advising an automobile manufacturer on a design decision that will be embedded in a consumer product affecting the public, Engineer A has an affirmative obligation under the principle of Completeness and Non-Selectivity in Advisory Opinions to disclose to the manufacturer that the harm-minimization recommendation reflects a specific moral philosophy, that alternative frameworks exist and yield different algorithmic outcomes, and that the selection among them is not a purely technical determination. Failure to make this disclosure would present the manufacturer with an incomplete picture of the decision it is actually making, impairing its ability to give informed consent to the embedded ethical framework and potentially exposing it to legal and reputational consequences it did not knowingly accept.

The interaction between the Autonomous System Moral Framework Transparency Obligation and the Regulatory Gap Safety Escalation Obligation - both activated by the absence of established national or industry standards governing autonomous vehicle harm-allocation ethics - produces a compounded disclosure duty that is stronger than either principle would generate in isolation. In the software testing context of BER Case 96-4, the regulatory gap triggered an obligation to flag the absence of standards as itself a safety concern and to recommend further study before deployment. Transposed to the autonomous vehicle harm-allocation context, that same gap-triggered escalation obligation combines with the transparency obligation to require Engineer A not only to recommend further study but also to affirmatively disclose to the automobile manufacturer that the harm-allocation recommendation rests on a specific moral framework - utilitarian harm minimization - rather than on a settled engineering standard. The Completeness and Non-Selectivity in Advisory Opinions principle reinforces this synthesis: because any recommendation Engineer A makes in a regulatory vacuum will necessarily reflect contestable ethical assumptions, selective silence about those assumptions would itself be a form of incomplete and potentially misleading professional advice. The case therefore teaches that regulatory vacuums do not relieve disclosure obligations; they intensify them, because the engineer's judgment substitutes for the absent standard and must therefore be rendered fully transparent.

The Board's conclusion that Engineer A must recommend minimizing harm to the least number of persons implicitly adopts a utilitarian ethical framework - specifically, an aggregate harm-minimization calculus - without acknowledging that this represents one among several defensible moral philosophies rather than a universally accepted engineering standard. A deontological framework, for instance, might prohibit the vehicle from actively redirecting harm toward any third party regardless of aggregate outcome, treating each person's life as inviolable rather than as a unit in a welfare sum. Because Engineer A is advising an automobile manufacturer on a design decision that will be embedded in a consumer product affecting the public, Engineer A has an affirmative obligation under the principle of Completeness and Non-Selectivity in Advisory Opinions to disclose to the manufacturer that the harm-minimization recommendation reflects a specific moral philosophy, that alternative frameworks exist and yield different algorithmic outcomes, and that the selection among them is not a purely technical determination. Failure to make this disclosure would present the manufacturer with an incomplete picture of the decision it is actually making, impairing its ability to give informed consent to the embedded ethical framework and potentially exposing it to legal and reputational consequences it did not knowingly accept.

In the absence of applicable regulatory or industry standards governing autonomous vehicle harm-allocation decision logic, Engineer A has an affirmative obligation to recommend that the automobile manufacturer publicly disclose the ethical framework embedded in the vehicle's operating system to prospective consumers before deployment. This obligation arises from the convergence of three independent sources: first, the Public Welfare Paramount principle, which requires that the public be protected not only from physical harm but from material deception about the nature of products that affect their safety; second, the Autonomous System Moral Framework Transparency Obligation, which recognizes that when an algorithm pre-commits to a harm-allocation outcome on behalf of a user who cannot intervene in real time, that user and affected third parties have a legitimate interest in knowing the decision logic governing their fate; and third, the regulatory standards vacuum itself, which - as the Board recognized analogously in BER Case 96-4 - heightens rather than relieves Engineer A's disclosure obligations precisely because no external regulatory body has yet stepped in to mandate transparency. The absence of a legal requirement to disclose does not extinguish the professional ethical duty to recommend disclosure. Engineer A's recommendation should therefore include not only the harm-minimization algorithm design but also a specific advisory that the manufacturer implement pre-sale consumer disclosure of the vehicle's decision logic as a condition of ethically responsible deployment.

The interaction between the Autonomous System Moral Framework Transparency Obligation and the Regulatory Gap Safety Escalation Obligation - both activated by the absence of established national or industry standards governing autonomous vehicle harm-allocation ethics - produces a compounded disclosure duty that is stronger than either principle would generate in isolation. In the software testing context of BER Case 96-4, the regulatory gap triggered an obligation to flag the absence of standards as itself a safety concern and to recommend further study before deployment. Transposed to the autonomous vehicle harm-allocation context, that same gap-triggered escalation obligation combines with the transparency obligation to require Engineer A not only to recommend further study but also to affirmatively disclose to the automobile manufacturer that the harm-allocation recommendation rests on a specific moral framework - utilitarian harm minimization - rather than on a settled engineering standard. The Completeness and Non-Selectivity in Advisory Opinions principle reinforces this synthesis: because any recommendation Engineer A makes in a regulatory vacuum will necessarily reflect contestable ethical assumptions, selective silence about those assumptions would itself be a form of incomplete and potentially misleading professional advice. The case therefore teaches that regulatory vacuums do not relieve disclosure obligations; they intensify them, because the engineer's judgment substitutes for the absent standard and must therefore be rendered fully transparent.

The Board's conclusion that Engineer A must recommend harm minimization for the least number of persons does not fully resolve what Engineer A's obligations become if the automobile manufacturer overrides that recommendation and elects to program the vehicle to prioritize passenger safety above third-party welfare. In that scenario, Engineer A's ethical obligations do not terminate upon delivery of the initial recommendation. Engineer A must first pursue graduated internal escalation within the risk assessment team and up the manufacturer's organizational hierarchy, clearly documenting the safety concern and its basis in the public welfare paramount principle. If internal escalation fails to produce a design that Engineer A can professionally certify as consistent with the obligation to hold paramount the safety, health, and welfare of the public - including pedestrians, cyclists, and motorcyclists who are third parties to the client relationship - Engineer A must consider whether continued participation in the project constitutes implicit endorsement of a harm-allocation algorithm that foreseeably causes fatal injury to third parties. At that threshold, refusal to certify the system or withdrawal from the engagement may be required. The consultant relationship does not diminish this obligation; the NSPE Code's public welfare paramount duty applies equally to consultants and employees, and the absence of a direct employment relationship does not reduce the enforceability of Engineer A's professional ethical duties.

If the automobile manufacturer, after receiving Engineer A's recommendation to minimize aggregate harm, decides to override that recommendation and program the vehicle to prioritize passenger safety above all others, Engineer A's ethical obligations do not terminate at the point of initial recommendation. Engineer A retains at minimum three residual obligations. First, under the principle of Graduated Internal Escalation Before External Reporting, Engineer A must formally document the disagreement and communicate to the manufacturer's decision-makers - in writing - that the passenger-priority algorithm creates a foreseeable risk of fatal harm to third parties that Engineer A regards as ethically unjustifiable, ensuring that the override decision is made with full awareness of its consequences rather than by default or inattention. Second, Engineer A must assess whether the resulting system design crosses the threshold from a debatable design choice into a design that Engineer A cannot in good conscience certify as safe for public deployment; if it does, Engineer A must decline to approve or certify the system under Code provision II.1.b., which prohibits approval of engineering documents not in conformity with sound engineering principles protective of public safety. Third, if internal escalation fails and Engineer A concludes that deployment of the passenger-priority algorithm poses an unreasonable risk of fatal harm to identifiable third-party classes - pedestrians, cyclists, motorcycle riders - Engineer A must evaluate whether external reporting obligations are triggered, recognizing that the NSPE Code's public welfare paramount obligation is not discharged merely by voicing concern internally when that concern is overridden and the harmful design proceeds.

The Board's conclusion that Engineer A must recommend harm minimization for the least number of persons does not fully resolve what Engineer A's obligations become if the automobile manufacturer overrides that recommendation and elects to program the vehicle to prioritize passenger safety above third-party welfare. In that scenario, Engineer A's ethical obligations do not terminate upon delivery of the initial recommendation. Engineer A must first pursue graduated internal escalation within the risk assessment team and up the manufacturer's organizational hierarchy, clearly documenting the safety concern and its basis in the public welfare paramount principle. If internal escalation fails to produce a design that Engineer A can professionally certify as consistent with the obligation to hold paramount the safety, health, and welfare of the public - including pedestrians, cyclists, and motorcyclists who are third parties to the client relationship - Engineer A must consider whether continued participation in the project constitutes implicit endorsement of a harm-allocation algorithm that foreseeably causes fatal injury to third parties. At that threshold, refusal to certify the system or withdrawal from the engagement may be required. The consultant relationship does not diminish this obligation; the NSPE Code's public welfare paramount duty applies equally to consultants and employees, and the absence of a direct employment relationship does not reduce the enforceability of Engineer A's professional ethical duties.

Engineer A's role as a consultant rather than a direct employee does not diminish the substantive scope of his ethical obligations under the NSPE Code, but it does affect the procedural mechanisms available to discharge them. The Code's public welfare paramount obligation applies with equal force to consultants and employees; Engineer A cannot invoke the consultant relationship as a basis for providing a narrower or more deferential safety assessment than an employee engineer would be required to provide. However, the consultant relationship does affect how far Engineer A must press concerns before his professional duty is satisfied in one specific respect: a consultant who has formally documented a safety concern, communicated it clearly to the client's responsible decision-makers, and been overruled has discharged the internal escalation component of his obligation more rapidly than an employee embedded in a hierarchical organization with multiple escalation tiers. The consultant's professional independence - which is itself a resource that the client engaged - means that Engineer A's obligation to provide an honest, complete, and unvarnished assessment of third-party harm risks is if anything stronger than that of an employee who might face internal organizational pressure to soften findings. Accordingly, Engineer A's consultant status heightens the independence and completeness obligations while compressing the internal escalation sequence, and does not create any basis for a reduced or qualified duty of care toward third-party public safety.

The interaction between the Autonomous System Moral Framework Transparency Obligation and the Regulatory Gap Safety Escalation Obligation - both activated by the absence of established national or industry standards governing autonomous vehicle harm-allocation ethics - produces a compounded disclosure duty that is stronger than either principle would generate in isolation. In the software testing context of BER Case 96-4, the regulatory gap triggered an obligation to flag the absence of standards as itself a safety concern and to recommend further study before deployment. Transposed to the autonomous vehicle harm-allocation context, that same gap-triggered escalation obligation combines with the transparency obligation to require Engineer A not only to recommend further study but also to affirmatively disclose to the automobile manufacturer that the harm-allocation recommendation rests on a specific moral framework - utilitarian harm minimization - rather than on a settled engineering standard. The Completeness and Non-Selectivity in Advisory Opinions principle reinforces this synthesis: because any recommendation Engineer A makes in a regulatory vacuum will necessarily reflect contestable ethical assumptions, selective silence about those assumptions would itself be a form of incomplete and potentially misleading professional advice. The case therefore teaches that regulatory vacuums do not relieve disclosure obligations; they intensify them, because the engineer's judgment substitutes for the absent standard and must therefore be rendered fully transparent.

The tension between the Faithful Agent Obligation - requiring Engineer A to serve the automobile manufacturer's interests - and the Third-Party Non-Client Welfare Consideration is real but resolvable within the NSPE Code's hierarchy of obligations. The Code does not treat these duties as co-equal: the public welfare paramount obligation is explicitly primary, and the faithful agent duty operates only within the ethical limits that the paramount obligation defines. This means that when the manufacturer's commercial interest in a passenger-protective algorithm conflicts with the safety of pedestrians, cyclists, and motorcycle riders, Engineer A is not required to balance these interests as if they were of equal weight. Instead, Engineer A must first satisfy the third-party safety obligation - by recommending the harm-minimization approach - and may then, within that constraint, seek to serve the manufacturer's interests by identifying technical solutions that minimize passenger harm within the harm-minimization framework. The faithful agent obligation does not authorize Engineer A to recommend a design that foreseeably causes fatal harm to third parties in order to protect the manufacturer's commercial position. What it does require is that Engineer A present the harm-minimization recommendation in a manner that is constructive, professionally grounded, and attentive to the manufacturer's legitimate interests in developing a commercially viable and legally defensible product - not that Engineer A suppress or soften the recommendation to accommodate those interests.

The tension between the Faithful Agent Obligation Within Ethical Limits and the Third-Party Non-Client Welfare Consideration is resolved in this case by treating the automobile manufacturer's commercial interest in a passenger-protective algorithm as categorically subordinate to the welfare of pedestrians, cyclists, and motorcyclists who bear the fatal risk of the vehicle's pre-committed harm-allocation logic. The Board's conclusion that Engineer A must recommend minimizing harm to the least number of persons effectively establishes a lexical ordering: Public Welfare Paramount operates as a side-constraint on the faithful agent role, not merely as one factor to be weighed against client interest. This means Engineer A's duty to serve the automobile manufacturer does not extend to endorsing an algorithm that systematically transfers lethal risk onto non-consenting third parties in order to protect paying passengers. The case teaches that when client interest and third-party safety are genuinely zero-sum - as they are in a pre-committed harm-allocation algorithm - the NSPE Code resolves the tension by collapsing the faithful agent role at the boundary where client service would require engineering complicity in foreseeable third-party fatalities.

From a deontological perspective, Engineer A has an obligation that is stronger than - and not fully captured by - the Board's utilitarian harm-minimization conclusion. The categorical imperative, applied to the autonomous vehicle harm-allocation problem, yields a distinct constraint: Engineer A must not recommend a design that treats any class of persons - whether passengers or third parties - as mere instruments for the benefit of another class. A passenger-priority algorithm that systematically redirects lethal force toward pedestrians treats pedestrians as means to passenger safety ends, which a Kantian analysis would prohibit regardless of aggregate welfare outcomes. Conversely, a pure harm-minimization algorithm that in specific scenarios sacrifices a single passenger to save multiple pedestrians may itself treat the passenger as a means to aggregate welfare ends. The deontological implication for Engineer A is not simply to recommend harm minimization, but to recommend that the design team explore whether any algorithm can be constructed that avoids pre-committing to the instrumental use of any person's life - for example, by designing for crash avoidance rather than crash outcome optimization, or by ensuring that the system's decision logic does not systematically disadvantage any identifiable class. Engineer A's obligation under this framework includes flagging to the manufacturer that the entire framing of the harm-allocation problem as a binary choice between passenger priority and aggregate minimization may itself embed morally problematic assumptions that warrant further study before deployment.

The Competing Public Goods Balancing principle - which acknowledges that vehicle passengers hold legitimate safety interests - does not neutralize the Public Welfare Paramount principle in this case; rather, the two principles interact to produce a qualified rather than absolute harm-minimization mandate. The Board's conclusion that Engineer A must recommend minimizing harm to the least number of persons implicitly acknowledges that passenger safety is a genuine public good, not merely a commercial preference, but treats aggregate harm reduction across all affected parties as the governing metric when those goods conflict. This resolution carries an important teaching: the Competing Public Goods Balancing principle functions as a corrective against naive utilitarian aggregation that would ignore passenger welfare entirely, while Public Welfare Paramount prevents that corrective from being weaponized to justify algorithms that predictably sacrifice a greater number of third-party lives to protect a smaller number of passengers. The net effect is that Engineer A's recommendation must be grounded in a harm-minimization calculus that counts all lives equally, resisting both pure passenger-priority logic and any framing that treats third-party lives as infinitely more valuable than passenger lives.

The interaction between the Autonomous System Moral Framework Transparency Obligation and the Regulatory Gap Safety Escalation Obligation - both activated by the absence of established national or industry standards governing autonomous vehicle harm-allocation ethics - produces a compounded disclosure duty that is stronger than either principle would generate in isolation. In the software testing context of BER Case 96-4, the regulatory gap triggered an obligation to flag the absence of standards as itself a safety concern and to recommend further study before deployment. Transposed to the autonomous vehicle harm-allocation context, that same gap-triggered escalation obligation combines with the transparency obligation to require Engineer A not only to recommend further study but also to affirmatively disclose to the automobile manufacturer that the harm-allocation recommendation rests on a specific moral framework - utilitarian harm minimization - rather than on a settled engineering standard. The Completeness and Non-Selectivity in Advisory Opinions principle reinforces this synthesis: because any recommendation Engineer A makes in a regulatory vacuum will necessarily reflect contestable ethical assumptions, selective silence about those assumptions would itself be a form of incomplete and potentially misleading professional advice. The case therefore teaches that regulatory vacuums do not relieve disclosure obligations; they intensify them, because the engineer's judgment substitutes for the absent standard and must therefore be rendered fully transparent.

From a deontological perspective, Engineer A has an obligation that is stronger than - and not fully captured by - the Board's utilitarian harm-minimization conclusion. The categorical imperative, applied to the autonomous vehicle harm-allocation problem, yields a distinct constraint: Engineer A must not recommend a design that treats any class of persons - whether passengers or third parties - as mere instruments for the benefit of another class. A passenger-priority algorithm that systematically redirects lethal force toward pedestrians treats pedestrians as means to passenger safety ends, which a Kantian analysis would prohibit regardless of aggregate welfare outcomes. Conversely, a pure harm-minimization algorithm that in specific scenarios sacrifices a single passenger to save multiple pedestrians may itself treat the passenger as a means to aggregate welfare ends. The deontological implication for Engineer A is not simply to recommend harm minimization, but to recommend that the design team explore whether any algorithm can be constructed that avoids pre-committing to the instrumental use of any person's life - for example, by designing for crash avoidance rather than crash outcome optimization, or by ensuring that the system's decision logic does not systematically disadvantage any identifiable class. Engineer A's obligation under this framework includes flagging to the manufacturer that the entire framing of the harm-allocation problem as a binary choice between passenger priority and aggregate minimization may itself embed morally problematic assumptions that warrant further study before deployment.

The Competing Public Goods Balancing principle - which acknowledges that vehicle passengers hold legitimate safety interests - does not neutralize the Public Welfare Paramount principle in this case; rather, the two principles interact to produce a qualified rather than absolute harm-minimization mandate. The Board's conclusion that Engineer A must recommend minimizing harm to the least number of persons implicitly acknowledges that passenger safety is a genuine public good, not merely a commercial preference, but treats aggregate harm reduction across all affected parties as the governing metric when those goods conflict. This resolution carries an important teaching: the Competing Public Goods Balancing principle functions as a corrective against naive utilitarian aggregation that would ignore passenger welfare entirely, while Public Welfare Paramount prevents that corrective from being weaponized to justify algorithms that predictably sacrifice a greater number of third-party lives to protect a smaller number of passengers. The net effect is that Engineer A's recommendation must be grounded in a harm-minimization calculus that counts all lives equally, resisting both pure passenger-priority logic and any framing that treats third-party lives as infinitely more valuable than passenger lives.

From a virtue ethics standpoint, Engineer A demonstrates the professional integrity and moral courage required of a virtuous engineer precisely by actively and unambiguously expressing concerns about harm-allocation algorithms within the risk assessment team, even when facing commercial pressure to prioritize passenger safety. Virtue ethics evaluates not only the content of Engineer A's recommendation but the manner and disposition with which it is made. A virtuous engineer in Engineer A's position would not merely file a technically correct recommendation and withdraw; he would engage substantively with the team's deliberations, articulate the moral stakes of the design decision in terms accessible to non-engineer stakeholders, and persist in raising concerns through appropriate channels if the initial recommendation is dismissed. The virtue of practical wisdom - phronesis - is particularly relevant here: it requires Engineer A to recognize that the harm-allocation problem is not purely technical, that the risk assessment team's composition and mandate may not be adequate to resolve the embedded ethical questions, and that recommending further interdisciplinary study before deployment is itself an expression of professional integrity rather than a failure to provide a definitive answer. A virtuous engineer does not manufacture false certainty about genuinely contested moral questions in order to satisfy a client's desire for a clean recommendation.

Beyond the Board's finding that Engineer A must recommend minimizing harm to the least number of persons, Engineer A bears an additional obligation to explicitly disclose to the automobile manufacturer that this recommendation is grounded in a utilitarian ethical framework rather than in any established regulatory or industry standard. Because no applicable national or industry standards governing autonomous vehicle harm-allocation decision logic currently exist, Engineer A cannot represent the harm-minimization recommendation as a technically mandated or universally accepted engineering norm. Presenting it as such would violate the completeness and non-selectivity obligation that governs Engineer A's advisory role. Engineer A must therefore clearly communicate to the automobile manufacturer that the recommendation reflects a specific moral philosophy - one that reasonable engineers and ethicists might contest - so that the manufacturer can make a genuinely informed deployment decision. This disclosure obligation is heightened, not relieved, by the regulatory standards vacuum, because the absence of external standards places the full burden of ethical transparency on Engineer A as the professional advisor.

If Engineer A had remained silent or provided only a partial assessment of third-party harm risks within the risk assessment team, the automobile manufacturer would not have had sufficient information to make an ethically informed deployment decision, and Engineer A's silence would have constituted a violation of both the faithful agent obligation and the public welfare paramount obligation. The faithful agent obligation requires Engineer A to provide the manufacturer with complete, accurate, and professionally grounded information relevant to the design decision - including information that is commercially inconvenient. Partial disclosure that omits the third-party harm implications of a passenger-priority algorithm would deprive the manufacturer of the ability to make an informed choice about the ethical and legal risks it is assuming. Simultaneously, Engineer A's silence would violate the public welfare paramount obligation by allowing a design to proceed toward deployment without the safety concerns having been formally raised, documented, and considered. The Code provision at III.1.b. - requiring engineers to advise clients when a project will not be successful - applies by analogy: a harm-allocation algorithm that foreseeably causes fatal harm to third parties in a predictable class of scenarios is not a successful engineering outcome, and Engineer A is obligated to say so. Silence in the face of a known, foreseeable, and serious public safety risk is not a neutral act under the NSPE Code; it is a breach of the engineer's professional duty.

If the automobile manufacturer had already established a firm design policy prioritizing passenger safety above all third-party considerations before Engineer A joined the risk assessment team, Engineer A's ethical obligations would shift materially - from recommendation toward escalation and, if necessary, refusal to certify. Under these circumstances, Engineer A's initial obligation to recommend the harm-minimization approach would remain, but its character would change: rather than being a prospective design input, it would function as a formal objection to an existing policy. Engineer A would be required to document that objection in writing, communicate it to the manufacturer's responsible decision-makers, and make clear that the existing passenger-priority policy creates foreseeable fatal risks to third parties that Engineer A regards as inconsistent with the public welfare paramount obligation. If the manufacturer declined to reconsider the policy after receiving this formal objection, Engineer A would face the question of whether to continue participating in the project. Continued participation in the design and certification of a system that Engineer A has formally identified as posing an unreasonable risk of fatal harm to third parties would be difficult to reconcile with the Code's prohibition on approving engineering documents not in conformity with sound engineering principles. Engineer A would therefore be obligated to decline to certify or approve the system, and to evaluate whether the severity and foreseeability of the third-party harm risk triggers any external reporting obligation under the public welfare paramount principle.

The Board's harm-minimization conclusion, while sound as a first-order ethical directive, does not adequately account for the possibility that a technically superior mitigation option - such as a sensor-based dynamic crash evaluation system capable of real-time scenario assessment rather than pre-committed algorithmic harm-allocation logic - could dissolve or substantially reduce the binary ethical dilemma between passenger safety and third-party harm minimization. Engineer A's obligation to explore additional technical mitigation options before accepting the dilemma as irreducible is itself an ethical duty, not merely a technical preference. Analogous to the reasoning in BER Case 96-4, where Engineer A was obligated to recommend further study and additional testing before deployment of safety-critical software, Engineer A in the present case must recommend that the risk assessment team investigate whether the harm-allocation decision can be made dynamically rather than pre-committed, thereby potentially achieving better outcomes for all parties across a wider range of crash scenarios. Recommending harm minimization without first exhausting technically feasible alternatives that could reduce the need for any pre-committed harm allocation would itself be an incomplete discharge of Engineer A's professional competence and public welfare obligations. If such alternatives are found to be technically infeasible, Engineer A must document that finding transparently so that the manufacturer's deployment decision is fully informed.

Had established national or industry standards governing autonomous vehicle harm-allocation decision logic existed at the time of Engineer A's assessment - analogous to the draft standards emerging in BER Case 96-4 - Engineer A's obligation to recommend further study before deployment would have been qualitatively different in character, though not necessarily stronger in absolute terms. The existence of applicable standards would have provided Engineer A with an external, professionally validated benchmark against which to evaluate the manufacturer's proposed algorithm, reducing the degree to which Engineer A's recommendation rested on Engineer A's individual ethical judgment. This would have made the recommendation more defensible, more actionable, and more likely to be accepted by the manufacturer. However, the absence of such standards does not weaken Engineer A's substantive obligation; it merely changes its epistemic basis. In the regulatory vacuum that actually exists, Engineer A's obligation to recommend further study is grounded in the recognition - itself drawn from the BER Case 96-4 analogy - that the absence of applicable standards is itself a safety-relevant fact that the manufacturer must be made aware of before deployment. The regulatory gap heightens the disclosure obligation and strengthens the case for recommending further interdisciplinary study, because it means that no external body has yet validated any harm-allocation approach as meeting a minimum standard of public safety. Engineer A's recommendation in the absence of standards must therefore be more explicitly provisional, more clearly flagged as reflecting one among several defensible approaches, and more strongly oriented toward recommending that deployment await the development of at least preliminary industry consensus.

The Board's harm-minimization conclusion, while sound as a first-order ethical directive, does not adequately account for the possibility that a technically superior mitigation option - such as a sensor-based dynamic crash evaluation system capable of real-time scenario assessment rather than pre-committed algorithmic harm-allocation logic - could dissolve or substantially reduce the binary ethical dilemma between passenger safety and third-party harm minimization. Engineer A's obligation to explore additional technical mitigation options before accepting the dilemma as irreducible is itself an ethical duty, not merely a technical preference. Analogous to the reasoning in BER Case 96-4, where Engineer A was obligated to recommend further study and additional testing before deployment of safety-critical software, Engineer A in the present case must recommend that the risk assessment team investigate whether the harm-allocation decision can be made dynamically rather than pre-committed, thereby potentially achieving better outcomes for all parties across a wider range of crash scenarios. Recommending harm minimization without first exhausting technically feasible alternatives that could reduce the need for any pre-committed harm allocation would itself be an incomplete discharge of Engineer A's professional competence and public welfare obligations. If such alternatives are found to be technically infeasible, Engineer A must document that finding transparently so that the manufacturer's deployment decision is fully informed.

If Engineer A had proposed and the team had successfully identified a technical mitigation option - such as a sensor-based system capable of dynamically evaluating crash scenarios in real time rather than relying on pre-committed algorithmic harm-allocation logic - the core ethical dilemma between passenger safety and third-party harm minimization would be substantially but not fully dissolved. A dynamic real-time evaluation system would eliminate the most ethically troubling feature of pre-committed harm-allocation logic: the systematic, categorical pre-assignment of fatal risk to identifiable classes of persons based on their mode of transportation rather than on the actual circumstances of a specific crash. However, Engineer A would retain significant residual ethical obligations even if such a system were technically feasible. First, Engineer A would be obligated to assess and disclose the reliability limitations of the dynamic evaluation system - including sensor failure modes, edge cases where real-time evaluation is impossible, and the possibility that the system's dynamic decisions might themselves embed implicit harm-allocation biases through the weighting of its input variables. Second, Engineer A would be obligated to recommend that the dynamic system's decision logic be made transparent to consumers and regulators, since the ethical concerns about algorithmic opacity do not disappear merely because the algorithm operates in real time rather than through pre-commitment. Third, Engineer A would be obligated to recommend that the dynamic system undergo further study and testing before deployment, since the novelty of the technology means that its real-world performance across the full range of crash scenarios cannot be validated through design analysis alone. The identification of a technical mitigation option reduces but does not eliminate Engineer A's public safety obligations.