Autonomous Vehicles Face Ethical Crossroads as Governance Lags Behind Innovation

By Yang Bowen, Yi Tong, and Jiang Guanghua
Beijing Research Center for Science of Science, Beijing 100089
Journal: Science and Technology Think Tank, 2021, Issue 1
DOI: 10.19881/j.cnki.1006-3676.2021.01.07

The global race to develop fully autonomous vehicles has reached a pivotal moment—not on the test track or in regulatory hearings, but in the moral frameworks that will govern how these machines make life-and-death decisions. As artificial intelligence (AI) continues to power next-generation mobility solutions, a growing chorus of experts warns that ethical governance is falling dangerously behind technological advancement. In China, this tension is especially acute, where rapid deployment of smart transportation systems clashes with an underdeveloped infrastructure for ethical oversight.

Autonomous driving technologies promise unprecedented safety, efficiency, and accessibility. Companies like Baidu, Pony.ai, and DeepRoute.ai have demonstrated Level 4 capabilities in urban environments across Beijing, Shanghai, and Shenzhen. State-backed initiatives under the “New Infrastructure” policy have accelerated the rollout of smart road networks embedded with 5G connectivity and AI-powered traffic management. Yet beneath this veneer of innovation lies a critical gap: who—human or machine—bears responsibility when an autonomous vehicle fails?

This question is not hypothetical. In recent years, multiple incidents involving advanced driver-assistance systems (ADAS) have raised alarms. While China has not yet reported a fatal crash directly attributable to full autonomy, near-misses and system malfunctions have exposed vulnerabilities in algorithmic decision-making. More fundamentally, they reveal a systemic absence of ethical protocols tailored to China’s unique mobility ecosystem—dense megacities, mixed traffic patterns, and a legal framework still adapting to AI-driven disruptions.

The ethical challenges extend beyond liability. Consider the “trolley problem” reimagined for Chinese streets: should an autonomous vehicle prioritize the safety of its passengers or swerve to avoid a jaywalking pedestrian? Such dilemmas are not philosophical abstractions but engineering choices encoded into software. Without transparent, publicly vetted ethical guidelines, these decisions are made in corporate boardrooms or research labs—often without input from the communities most affected.

China’s approach to AI governance has been proactive in principle but fragmented in practice. In 2019, the Ministry of Science and Technology established the National New-Generation Artificial Intelligence Governance Professional Committee, signaling high-level recognition of ethical risks. That same year, the committee released the “Governance Principles for New-Generation Artificial Intelligence: Developing Responsible AI,” which affirmed core values including human well-being, fairness, and accountability. However, these principles remain largely aspirational. They lack binding force, enforcement mechanisms, or sector-specific implementation standards—particularly for automotive AI.

Regulatory efforts have been similarly cautious. The 2018 “Administrative Specifications for Road Testing of Intelligent Connected Vehicles (Trial),” jointly issued by the Ministry of Industry and Information Technology, the Ministry of Public Security, and the Ministry of Transport, focused primarily on technical safety and operational procedures. It required test drivers to remain ready to take control and mandated accident reporting—but said little about ethical design or moral decision architecture. Subsequent provincial-level guidelines in Guangdong, Zhejiang, and Beijing mirrored this technical bias, sidestepping deeper normative questions.

This regulatory gap contrasts sharply with international developments. The European Union’s 2019 “Ethics Guidelines for Trustworthy AI” explicitly demands “human oversight” and “societal and environmental well-being,” principles that could directly inform autonomous vehicle design. Germany’s 2017 Ethics Commission on Automated and Connected Driving went further, issuing 20 rules that prohibit discrimination in algorithmic decisions and mandate data transparency—making it the world’s first national framework to legally enshrine ethical AI in transport.

In the United States, while federal legislation remains stalled, industry self-regulation has filled part of the void. Companies like Waymo, Cruise, and Tesla have published AI ethics charters, and consortia such as the Partnership on AI—co-founded by Google, Microsoft, Amazon, and IBM—develop shared standards for responsible deployment. Though imperfect, these efforts foster public dialogue and create de facto benchmarks for ethical performance.

China lacks a comparable ecosystem of cross-sector collaboration. Academic institutions like Tsinghua University and the Chinese Academy of Sciences have produced thoughtful papers on AI ethics, and the Beijing Academy of Artificial Intelligence issued the “Beijing AI Principles” in 2019, followed by a child-focused addendum in 2020. Yet these remain voluntary frameworks with limited traction in industrial practice. Meanwhile, public awareness of AI ethics remains low. A 2018 China Association for Science and Technology survey found that only 5% of researchers reported deep familiarity with ethical norms beyond research integrity—a troubling indicator for a field as socially embedded as autonomous driving.

The consequences of this oversight could be profound. As AI reshapes labor markets, millions of professional drivers—truckers, taxi operators, delivery personnel—face displacement. In Kunshan alone, Foxconn’s automated factories have eliminated 60,000 jobs, offering a glimpse of the transport sector’s future. Without proactive policies for retraining and social protection, AI-driven unemployment could exacerbate inequality, fueling public distrust in technological progress.

Moreover, data privacy remains a critical vulnerability. Autonomous vehicles collect terabytes of data daily—geolocation, behavioral patterns, even biometric information from cabin sensors. While China’s 2021 Personal Information Protection Law (PIPL) establishes baseline privacy rights, enforcement in the automotive context is nascent. A 2019 scandal involving the “ZAO” deepfake app, which harvested facial data under opaque consent terms, illustrates how easily AI systems can exploit regulatory gray zones. Connected vehicles present even greater risks: a compromised fleet could enable mass surveillance or targeted manipulation.

To address these challenges, a multi-pronged governance strategy is urgently needed. First, China must establish a dedicated national AI ethics oversight body with statutory authority over high-risk sectors like autonomous transport. This entity should include not only technologists and legal scholars but also ethicists, civil society representatives, and industry practitioners to ensure balanced deliberation.

Second, ethical principles must be operationalized into technical standards. For instance, the China Electronics Standardization Institute could develop mandatory certification protocols requiring all autonomous systems to undergo ethical impact assessments—similar to environmental impact reviews for infrastructure projects. These assessments would evaluate fairness in decision algorithms, robustness against bias, and transparency in data usage.

Third, legal clarity on liability is essential. China should amend its Civil Code or introduce specific AI liability legislation to define responsibility chains: from software developers and sensor manufacturers to fleet operators and end users. Drawing on Germany’s model, the law could presume operator liability unless the manufacturer proves the system functioned as designed—a reversal that incentivizes rigorous safety engineering.

Fourth, public engagement must be institutionalized. Forums for citizen deliberation—town halls, citizen juries, participatory design workshops—can help democratize ethical choices in AI. When millions of vehicles will soon make split-second moral decisions on public roads, those decisions cannot be left to engineers alone.

Finally, China should leverage its technological leadership to shape global norms. By championing inclusive, human-centered AI governance at forums like the G20, OECD, and United Nations, China can position itself not just as an innovator but as a responsible steward of emerging technologies. Participation in international standard-setting bodies—such as ISO/IEC JTC 1/SC 42 on AI—offers a platform to export ethical best practices while learning from global peers.

The window for proactive governance is narrowing. Every autonomous vehicle deployed without an ethical compass increases the risk of public backlash, regulatory overreach, or catastrophic failure. The goal is not to stifle innovation but to channel it toward outcomes that align with societal values: safety, dignity, fairness, and sustainability.

China’s ambition to lead in intelligent mobility is clear. But true leadership requires more than technical prowess—it demands moral foresight. As the nation accelerates toward a driverless future, the road ahead must be paved not only with silicon and steel but with wisdom, accountability, and a steadfast commitment to human dignity.

The autonomous revolution will not be judged solely by how fast vehicles drive, but by how wisely they decide. China now stands at that ethical crossroads. The choices it makes today will echo through streets, courts, and conscience for generations.

—