ARTIFICIAL INTELLIGENCE, LAW AND GENDER EQUALITY
An Examination of the Ethical, Legal, and Social Dimensions of Technological Transformation
Nazan MOROGLU, LLM
- Introduction and the Artificial Intelligence Revolution
Artificial intelligence (AI), in its broadest sense, refers to the development and use of systems that mimic human cognitive processes—such as learning, reasoning, problem-solving, and decision-making through machines. Today, this technology permeates every aspect of life, from healthcare to education, from the economy to security, and is a key driver of digitalization. Systems developed by leading companies such as OpenAI, Google, Microsoft, Anthropic, and DeepSeek are laying the foundation for a new world order. This profound transformation can be described as an “AI revolution.” AI systems help organizations make faster, more accurate, and more effective decisions by processing big data. This technology, which increases efficiency in business processes and saves both time and money,
has become an indispensable component of modern businesses.
- Risks, Ethical Principles, and Striking a Balance
In addition to adopting and effectively using technology, it is essential to take proactive measures against the risks that this process may entail. While artificial intelligence offers countless benefits, we must not lose sight of the fact that the ultimate goal is “human benefit” and “for the sake of humanity.” Accordingly, a delicate balance must be struck between the advantages offered by AI and ethical principles, fundamental human rights, and the rule of law.
When designing AI, “security”—the most fundamental need that humans have sought since the dawn of humanity—must be prioritized; the principles of accuracy, transparency, and accountability must be carefully upheld. Uncontrolled use without the necessary oversight mechanisms and precautions could lead not only to material damage but also to irreparable losses of trust and human rights violations in social life. For this reason, at every stage—from the uploading of data into the system to the writing of algorithms and their delivery to the end user—we must not merely adhere to abstract ethical principles; rather, these principles must be anchored in legal safeguards.
- Artificial Intelligence from a Legal Perspective and an Interdisciplinary Approach
As societal needs evolve and change, legal rules must also keep pace with this evolution. However, the gap between the pace of technological advancement and the pace of legal regulation is widening with each passing day. It is of critical importance that legislation clarifying issues of accountability and legal liability in artificial intelligence systems be enacted without delay.
Data processing, algorithm design, and the establishment of standards for system use transcend the boundaries of a single discipline. To ensure that legal regulations in this field are grounded on a sound foundation, an interdisciplinary commission must be established. Collaboration among the following stakeholders is essential within this commission:
Legal Experts: Protection of rights, establishment of a liability regime, and definition of the legal framework.
Engineers and Technical Experts: Algorithmic transparency, system design, and oversight of the technical infrastructure.
Ethics Experts: Integration of universal human rights and moral values into the system.
Economists and Competition Experts: Digital markets, data monopolization, and the protection of economic balances.
Industry Representatives and Local Governments: Sharing field experience from the education, healthcare, finance, and justice sectors.
With contributions from all these components, the legal framework that will take shape is expected to become institutionalized in the near future as an independent discipline within the legal world under the name “Artificial Intelligence Law.”
- Current Status and Needs in Turkish Law
There is currently no independent and comprehensive legal framework specific to artificial intelligence in the Turkish legal system. However, the provisions in our existing legislation (the Constitution, the Law on the Protection of Personal Data [KVKK], the Turkish Code of Obligations [TBK], the Turkish Penal Code [TCK], the Labor Code, and the Law on the Protection of Consumers) contain provisions that, by their very nature, can be applied to algorithms and data processing procedures.
However, in current practice, significant legal gaps and deficiencies are observed in the context of transparency, full accountability, and combating algorithmic discrimination. In this regard, it has become an urgent necessity for Turkey to develop a risk-based and dynamic national artificial intelligence law similar to the Artificial Intelligence Act (AI Act) enacted by the European Union. The Bill, submitted to the Grand National Assembly of Turkey (TBMM) in July 2025 and proposing the introduction of criminal and legal sanctions for artificial intelligence applications, as well as amendments to various laws in this regard, is still awaiting inclusion on the legislative body’s agenda.
A Historical Perspective: Ord. Prof. Dr. Cahit Arf (1959)
The origins of discussions in the field of artificial intelligence in our country date back quite a long time. As a concrete example of the Republic’s educational vision based on reason, science, and the enlightenment of the people, in 1959, Ord. Prof. Dr. Cahit Arf delivered a visionary speech titled “Can Machines Think, and How Can They Think?” at the Atatürk University Public Conferences in Erzurum. Prof. Arf concluded his speech with the following words: “Let me conclude my remarks with a brief comparison between these machines and the human brain: Machines can perform certain tasks much more quickly than the human brain. While the human brain is capable of self-improvement, a machine remains as it was originally built. However, it is possible to design a machine capable of self-improvement.” This observation is an early foreshadowing of today’s machine learning technologies.
- The Relationship Between Artificial Intelligence and Gender Equality
Although artificial intelligence systems are perceived as neutral and objective, the datasets they are trained on and the methods by which they are designed pose a risk of perpetuating—or even amplifying—existing societal biases. The experiences of global technology giants provide concrete evidence of this situation:
Case Study 1: The Amazon Hiring Algorithm Example
Development Process: In 2014, Amazon began developing an AI system to automate and accelerate its hiring processes.
Problem and Bias: The system was trained using applications and resumes submitted to the company over the past 10 years. Due to the historically high proportion of male employment in the technology sector, the algorithm established a flawed correlation suggesting that “successful candidates are male.” The system systematically lowered the scores of candidates whose resumes contained phrases including the word “woman”—such as “president of a women’s chess club”—or who graduated from universities that offered programs exclusively for women.
Result: Despite the engineers’ algorithmic interventions, the system’s complete impartiality could not be guaranteed, and the project was completely canceled in 2018. This case holds an important place in the literature for demonstrating the depth of historical bias.
Case Study 2: Google and Data Diversity Principles
Issue: In the image recognition and language models developed by Google, occupational gender bias has been observed as a result of historical data. For example, it has been found that in the systems’ translation tools or image matching features, the concept of “doctor” is predominantly associated with male profiles, while the concept of “nurse” is associated with female profiles (or related pronouns).
Solution and AI Principles: To address such structural issues, Google has published a binding declaration titled “AI Principles.” Under this framework, algorithms have been optimized by being retrained using inclusive, highly diverse datasets that encompass different genders, ethnic backgrounds, and age groups.
Digital Violence and the GREVIO Recommendation
Another critical threat posed by advancing technology in the context of gender equality is digital violence against women and girls. To prevent this new form of violence and effectively protect victims, the GREVIO (Expert Group on Action against Violence against Women and Domestic Violence) published General Recommendation No. 1, “The Digital Dimension of Violence against Women,” on October 20, 2021. This international document imposes significant obligations on states regarding the protection of women on digital platforms.
- Conclusion and General Assessment
Artificial intelligence is not, by its very nature, inherently anti-gender or discriminatory. If algorithms are designed from the outset with an egalitarian perspective and the datasets fed into the system are free of gender-based discrimination, AI can become a powerful tool capable of identifying and reporting social inequalities and accelerating progress in this area.
However, if the process is left uncontrolled and unmonitored, current AI technologies will institutionalize and entrench existing patriarchal gender patterns. As a result, we face the risk that the multifaceted rights struggle waged for many years—from legislation to the elimination of discrimination against women and the realization of equal rights—will now have to be waged against AI algorithms. A just society of the future can only be built through equitable algorithms and a robust AI Law.
BIBLIOGRAPHY AND REFERENCES
- Moroğlu, N. “Adapting Artificial Intelligence to Legal Processes in 10 Steps.” Accessed at: https://nazanmoroglu.com/10-adimda-yapay-zekayi-hukuksal-sureclere-adapte-etmek/
- Moroğlu, N. “GREVIO General Recommendation No. 1 on Digital Violence Against Women.” Available at: https://nazanmoroglu.com/kadinlara-yonelik-dijital-siddet-hakkinda-grevio-1-no-lu-geneltavsiye/
- Prof. Dr. Cahit Arf, “Can Machines Think, and How Do They Think?”, Atatürk University Public Lectures, Erzurum, 1959. Archive Access Date: December 12, 2025 (Gmail Archive / cahit arf)
