AI in Human Rights Protection: A New Horizon for Supply Chains

Major companies worldwide face immense pressure and complexity in overseeing their global supply chains, especially concerning human rights abuses (e.g., child labor in cobalt mining). Such incidents can severely damage a company’s reputation. While concerns exist about AI contributing to human rights violations, it also offers significant potential to improve human rights due diligence.

Challenges of Traditional Due Diligence

Traditional methods like supplier surveys are often ineffective as disclosure requests are ignored, and the multi-tiered nature of supply chains creates insurmountable analytical challenges, leading to a “paralysis of analysis”.

AI’s Transformative Contribution

Artificial intelligence is an “enormous boon” for companies, helping them “see, focus and act” on human rights risks. AI’s key contributions include:

• Unifying vast amounts of data at scale, such as procurement and shipping records, to map entire global supply chains.
• Identifying potential violations by flagging points where human rights violation probabilities are higher or irregular patterns in supplier data indicate hidden abuses.
• “Noise suppression”, cutting through irrelevant data to focus on direct connections to finished goods.
• Enabling more strategic auditing and aiding responsible sourcing teams in building business cases for interventions by conducting cost-benefit analyses on switching suppliers.
• Providing greater scalability and quicker alerts when supply chain partners are linked to violations.

Companies like Altana and FRDM.ai are actively developing these AI tools, and initiatives like the Global Data Partnership Against Forced Labour leverage AI for insights into forced labor prevalence.

Regulatory Landscape:

Growing regulatory pressure, including the U.S. Dodd-Frank Act (2010), the European Commission’s Corporate Sustainability Due Diligence Directive, and proposed legislation in South Korea, along with “soft law” like the U.N. Guiding Principles, are compelling companies to enhance human rights due diligence in their supply chains. The U.S. Uyghur Forced Labor Prevention Act (2022) further exemplifies this trend, leading to impoundment of goods.

Limitations and Future Outlook:

Despite considerable optimism, AI is not a “silver bullet”. Its effectiveness is fundamentally limited by data quality and availability, especially in regions where abuses are less likely to be digitally reported. Critically, AI cannot replace the essential human element of engaging with stakeholders and “talking to rights holders, listening to their perspectives”. While AI tools will play an increasingly vital role in identifying risks, whether this leads to actual on-the-ground change and stops human rights abuses remains a broader challenge beyond just AI solutions.

See Reference: How AI is helping companies zero in on human rights risks