The New EU AI Act and Türkiye’s AI Bill

Türkiye’s AI Bill:

The bill stipulates mandatory principles to ensure AI’s safe, ethical and fair usage. It includes guidelines to make sure human rights are respected within the rapid development and accelerating practice of AI. The proposal states that the bill is applicable to; 

• Providers
• Deployers
• Importers
• Distributors
• Individuals affected by AI systems

Although the EU AI Act and Türkiye’s bill both aim to uphold human rights through introducing a risk-based approach on AI operations, differences may be listed concerning definitions, high risk classification and penalties. 

Human Rights Impacts of the Bill

• Upholding the right to security by ensuring reliability, durability and safety of the systems and identifying potential dangers and preventing them in advance.  
• Establishing transparency in AI decision-making processes by providing comprehensive explanations on functioning of AI systems and criterias upheld in decision-making. 

• Acknowledging prohibition of discrimination to provide fair AI systems that do not automatically generate responses based on factors such as gender, age, race, religion, ethnicity and treat all individuals equally. 

• Attaining accountability to those responsible for adverse human rights impacts to address the right to effective remedy for affected individuals and communities.

• Protection of the right to private life through establishing necessary measures to ensure confidentiality and delimit the usage of personal data. 

Rise of AI from the Perspective of Business and Human Rights

Utilisation of AI systems has emerged as rapid and innovative technological development following the endorsement of UN Guiding Principles on Business and Human Rights (UNGPs). 

The implementation of the UNGPs, which encompass three pillars of the State’s duty to protect human rights, the corporate responsibility to respect human rights, and the provision of access to remedies for victims, must be tailored to the specific characteristics of AI systems due to the unpredictability and ambiguity of traditional roles within the supply chain, where the functions of suppliers, producers, and consumers often overlap and intertwine.

Integrating Human Rights and Environmental Due Diligence and Stakeholder Engagement

1. To fulfil the state’s duty to protect human rights under UNGPs, states should ensure companies carry out a human rights and environmental due diligence (HREDD) process as part of their obligation to respect human rights by creating the necessary legislation. This process includes a system for risk management and reporting, which involves conducting regular risk assessments and implementing grievance mechanisms.
2. Furthermore, as stakeholder engagement is vital in HREDD, states should have effective open communication with AI operators and other stakeholders  such as NGOs and human rights defenders while creating new legislation to evaluate and mitigate AI’s impacts on individuals and communities.