AI and ethics最新文献

Artificial intelligence is increasingly used to screen job applicants, yet empirical studies show that algorithmic hiring systems can reproduce racial, gender, and intersectional disparities. Existing mitigation strategies frequently treat these disparities as technical problems—addressed through data adjustments, model tuning, or compliance metrics—while insufficiently engaging with the broader ethical implications of discriminatory outcomes. This paper examines algorithmic bias in hiring through established ethical concepts of justice, capability development, and recognition to explain why technical and regulatory approaches alone prove inadequate. The analysis synthesizes empirical evidence of algorithmic discrimination with normative frameworks to identify structural limitations in prevailing fairness interventions. Findings reveal that while technical methods can reduce measurable disparities, they often fail to address deeper concerns: restricted access to meaningful opportunities, the devaluation of marginalized groups, and violations of procedural fairness. Regulatory frameworks similarly prioritize compliance over substantive commitments, contributing to “bias washing” where organizations meet formal requirements without achieving genuine equity. To address these limitations, the paper proposes an “ethical by design” governance model that integrates normative principles with technical implementation, guiding organizations toward hiring systems that promote equitable opportunity distribution and uphold candidate dignity.

Artificial intelligence (AI) is accelerating discovery timelines in synthetic biology, expanding opportunities for therapeutic breakthroughs, sustainable bio-manufacturing, and rapid response to health and environmental challenges. At the same time, this acceleration shifts biosecurity risks from physical materials toward a broader socio-technical landscape involving models, datasets, and distributed automation. This literature review synthesizes evidence from 119 peer-reviewed articles published between January 2015 and August 2025, focusing on biosecurity risks, dual-use concerns, and governance responses at the AI–synthetic-biology interface. Findings indicate that AI systems consistently increase design throughput and lower expertise barriers, enabling faster medical and industrial innovation but also heightening risks of repurposing for harmful molecules or genetic sequences. Existing governance remains fragmented: biosafety regimes emphasize physical agents and laboratories, while AI governance frameworks focus on privacy and fairness—leaving critical blind spots for biological misuse scenarios. Mitigation measures identified in the literature converge on layered controls, including risk-tiered access to high-capability models, systematic red-teaming prior to release, strengthened DNA-synthesis screening (including short fragments), audit logging, secure data infrastructures, and international capacity-building. Evaluation gaps persist, and harmonized metrics—such as synthesis-screening coverage, red-team testing frequency, accredited biofoundries, and early-warning lead times—are recommended for systematic monitoring of governance effectiveness. Addressing these challenges is essential to ensure that AI-enabled synthetic biology advances responsibly, balancing its transformative potential for health and sustainability with global biosecurity.

Introduction

The rapid advancement of artificial intelligence raises profound technical, ethical and philosophical questions: Can machines attain human-like dignity, and if so, how do we decide that? This article presents “Immanuel”, a blueprint for an intelligent machine designed to meet the criteria of human dignity as defined in international human rights frameworks.

Methods

Immanuel was developed by designing technical representations of human traits as described in typical textbooks about physiology, psychiatry, and psychology. In order to prove that Immanuel does indeed have human dignity, the individual components of dignity (as described earlier) were checked individually.

Results

Immanuel’s design integrates perception, self-modeling (an “I” module), and ethical valuation. All features are implementable with current technology and enable human-like moral feelings, learning, empathy, and autonomous decision-making. Systematic comparison confirms that Immanuel fulfills all criteria for human dignity.

Discussion

As of today, it is possible to build machines with human dignity. Therefore, the discussion about ethical implications must be held urgently.

Integrating Artificial Intelligence (AI) tools, particularly generative AI (GenAI), in higher education is reshaping assessment practices, presenting both challenges and opportunities. While these tools enhance learning, they also raise concerns about academic integrity and the authenticity of student work. Traditional assessments, such as essays and take-home assignments, are increasingly susceptible to AI-assisted plagiarism, necessitating a re-evaluation of assessment strategies. This systematic review, guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework, examines educators' challenges in assessing student learning in AI-infused environments. Using Scopus, IEEE Xplore, and ScienceDirect, we identified relevant literature highlighting concerns about originality, critical thinking evaluation, and the quality of student work. Findings underscore the need for AI-resistant, process-based assessments, such as oral exams and multi-stage evaluations, to uphold academic integrity. The study advocates for institutional AI policies and digital literacy programs to promote ethical AI use and mitigate academic misconduct. Additionally, it emphasises a balanced human-AI collaboration in assessments, ensuring that AI enhances rather than replaces student effort. Addressing these challenges can reduce academic misconduct cases, allowing educators to focus on fostering meaningful learning experiences and sustainable educational outcomes.

This study explores the transformative impact of conversational artificial intelligence (AI) on customer support systems across various industries. Through a systematic qualitative literature review of 15 empirical studies and the analysis of 10 contemporary abstracts, the research identifies key operational, experiential, and ethical dimensions associated with AI-driven service delivery. Findings reveal that conversational AI significantly enhances efficiency, scalability, and customer accessibility, especially during crisis periods like the COVID-19 pandemic. However, challenges such as limited emotional intelligence, ethical concerns over data privacy and bias, and lack of transparency persist. The review highlights that customer satisfaction and trust depend heavily on the system's communication quality, anthropomorphic design, and ability to escalate complex issues to human agents. Hybrid models, combining AI automation with human oversight, consistently produce the most favorable outcomes. Sector-specific variations underscore the importance of context-aware AI deployment strategies. The study emphasizes the urgent need for ethical frameworks, regulatory standards, and user-centered design in the development and implementation of conversational AI technologies. Future research should focus on longitudinal, real-world evaluations to address gaps in user experience, fairness, and technological accountability. This paper contributes to the understanding of how AI can be sustainably integrated into customer service, aligning innovation with human-centric values.

The integration of Artificial Intelligence (AI) in education has caused increasing concerns over principles of ethics, including transparency, fairness, privacy, and accountability. Although several literature reviews have investigated these issues separately, there is a lack of consistency among research, leading to fragmented terminology and conceptual overlap. The present study aims to perform a meta-review of current literature studies to identify, categorize, and consolidate the ethical considerations associated with AI usage in education. This study systematically analyzed 13 peer-reviewed literature review articles published from 2021 to 2025, adhering to the PRISMA 2020 framework. Articles were chosen according to rigorous inclusion criteria, highlighting reviews that examined ethical issues in AI applications within various educational contexts. A qualitative thematic synthesis was employed to extract and categorize ethical concepts into broad themes. The research revealed seven key ethical themes: Transparency and Accountability, Data Protection and Security, Fairness and Non-Discrimination, Human-Centric AI design, Ethical Responsibility and Governance, Moral Principles in AI, and Technical Integrity and Robustness. These findings provide a unified thematic framework that encompasses both fundamental ethics and contemporary challenges, including those presented by generative AI. The study identifies gaps in the literature by providing a systematic ethical framework and suggests future research directions that include empirical validation, cross-cultural investigations, and an examination of generative AI risks. This review theoretically aids in establishing a cohesive ethical foundation to guide future models and governance frameworks for responsible AI in education.

In recent months, there has been a noticeable shift toward AI deregulation and soft law, particularly in the U.S.—exemplified by Trump’s revocation of Biden’s executive order—and, to a lesser extent, in the E.U., as seen in the adoption of a voluntary code of conduct for providers of general-purpose AI. In contrast to these political trends, this paper argues that the significant risks posed by AI technologies call for stronger and more effective governance mechanisms. It presents ten key reasons why robust AI regulation is essential: the impact of AI on labor markets and social equity (i.e., the global/digital divide); (agentic) AI biases and discrimination; infringements on privacy; the amplification of harmful online content, including dangerous and hate speech and disinformation; business and human rights concerns; environmental consequences; the growing politico-economic power of big tech companies; threats to democracy and the rule of law; and the military implications of emerging AI technologies. Finally, the paper explores potential pathways for future AI regulation (soft vs. hard law and constitutional AI vs. AI guardrails) and proposes strategies to enhance existing regulatory frameworks, such as the E.U.’s AI Act and competition policy.

This paper shows that superintelligent AI (ASI) poses a significant risk to meaning in human life by relying on Susan Wolf’s conception, according to which meaningful lives are lives of active engagement in projects of worth. The paper argues that ASI makes it less likely for humans to lead meaningful lives by reducing the possibility of human contribution and active engagement in what are deemed to be some of the most worthwhile projects of human life. The paper also criticizes Nick Bostrom’s and John Danaher’s views on how meaning may be retained in spite of this.

This paper introduces the Capability-Sensitive Framework (CSF), a formal method for auditing AI systems through the lens of capability approach. CSF specifies two normative guardrails: a capability floor, which ensures no individual is pushed below thresholds for essential freedoms, and a life-plan ceiling, which guarantees that people retain viable paths toward their meaningful goals. These constraints are operationalized via two metrics, the Capability-Coverage Ratio (CCR) and Life-Plan Alignment Score (LAS), evaluated at both individual and subgroup levels. A typology of positive and negative modalities situates benefits and harms in capability terms, distinguishing ethical outcomes from systemic risks and clarifying category boundaries. CSF foregrounds human flourishing, ethical sufficiency and human-centric values, enabling actionable, model-agnostic guidance for high-stakes sociotechnical decisions. This work bridges normative political philosophy with formal auditing practices for AI systems, offering a rigorous foundation for capability-aware AI governance.

The opacity of AI systems poses a fundamental epistemic challenge: how can we justifiably trust systems whose decision processes are inscrutable? Rather than relying on inherently limited explainability, I propose an institutional epistemology of warrant. Drawing on speech act theory and social epistemology, I argue that institutional sources that constitute (rather than merely describe) reality can ground epistemic warrant without algorithmic transparency. The framework addresses both retrieval scenarios (existing constitutive sources are cited and verified) and generative scenarios (AI outputs acquire force only after institutional validation). It shifts attention from internal mechanisms to verifiable linkages with authoritative institutional frameworks. Despite important limitations (risks of fabricated constitutive sources, vulnerability to institutional decline, and ethical concerns about algorithmically originated normative content) the account offers a philosophically grounded alternative where comprehensibility is unattainable. The framework is intended for domains where constitutive authority enjoys recognized legitimacy, presupposing rather than creating institutional trust. The framework presupposes existing institutional trust; yet by making institutional warrant more visible and auditable, it also has the potential to reinforce that trust ex post.