Artificial Intelligence Looks Toward Drafting the Laws of the Future

Introduction

The world today is experiencing an unprecedented digital revolution. Just as the Industrial Revolution compelled legislators to develop laws governing labor, industrial property, and liability for machinery, the artificial intelligence revolution is forcing a reconsideration of a number of traditional legal concepts, from civil liability and data protection to intellectual property and transparency in decision-making.

Artificial intelligence is no longer merely a technical tool for improving productivity. It has become capable of analyzing vast quantities of legislation, judgments, and data, identifying inconsistencies and deficiencies, comparing legislative alternatives, and even proposing initial drafts of legal provisions. This raises a question for the future: Can artificial intelligence participate in the law-making process itself?

Dr. Mostafa El Rouby previously addressed the impact of modern technological developments on legal rules in earlier articles, emphasizing that the speed of innovation now requires legislators to move beyond merely addressing problems after they arise toward building legislation capable of accommodating development before it turns into a crisis.

From this perspective, the issue is not about replacing the human legislator with a machine, but rather about how artificial intelligence can be used to improve the quality of legislation while keeping final decision-making and responsibility in human hands.

The Digital Revolution and Legislative Challenges

Artificial intelligence systems raise legal issues for which traditional rules do not always provide direct answers. Who bears responsibility if an autonomous or semi-autonomous system causes harm? What are the acceptable limits on the use of algorithms in recruitment, credit decisions, or healthcare? How should personal data used to train models be protected? And who owns content produced with the assistance of artificial intelligence?

The issue becomes even more complex in deep-learning systems, where it can sometimes be difficult to explain the path that led the model to a particular result, a problem commonly known as the “black box”.

Addressing this problem does not mean that companies must always disclose their source code or proprietary technical secrets to the public, as this could conflict with intellectual property rights, trade secrets, and cybersecurity. Rather, the challenge lies in establishing an appropriate level of documentation, transparency, explainability, and auditability proportionate to the seriousness of the decision made by the system.

It would not be logical to subject an application that recommends music to users to the same restrictions imposed on a system that assists in making decisions relating to employment, credit, or public services.

The European Union and the Artificial Intelligence Act

The European Union Artificial Intelligence Act (EU AI Act) represents the most prominent comprehensive legislative experiment in this field. The Act entered into force on 1 August 2024, and its provisions began to apply in stages. Most of the Act became applicable as of 2 August 2026, with specific transitional periods for certain obligations.

The European system is based primarily on a risk-based approach. Certain practices are considered unacceptable and prohibited by law, high-risk systems are subject to stricter obligations, and certain cases require transparency obligations, while low-risk uses are not subject to the same degree of regulation.

Rules prohibiting certain artificial intelligence practices and requirements relating to AI literacy began to apply in February 2025, while governance rules and obligations relating to general-purpose artificial intelligence models became effective in August 2025.

What Does “High-Risk Artificial Intelligence” Mean?

European regulation does not look solely at the technology itself, but rather at the context in which it is used and the rights that may be affected by its decisions.

Systems used in certain sensitive areas, such as recruitment, education, critical infrastructure, and certain essential services, may be subject to stricter controls, including risk management, data quality, documentation, record-keeping, cybersecurity, and human oversight.

This legislative philosophy is important for countries preparing their own laws because it avoids two opposing mistakes: leaving technology completely unregulated, or subjecting all applications to strict restrictions that hinder innovation.

Human Oversight Remains at the Core of the Legal System

One of the most prominent principles reflected in modern legislation is the need to preserve the role of humans in high-impact decisions.

Artificial intelligence may be able to analyze thousands of facts and documents within seconds, but it does not bear legal or ethical responsibility in the same way as a human being or legal entity that decided to use it.

This highlights the importance of Human Oversight, meaning the presence of a qualified person capable of understanding the system’s limitations, reviewing its output, and intervening or suspending its use when a risk or error appears.

The closer a decision comes to affecting individuals’ fundamental rights, the more important it becomes to ensure that the human being does not become merely an automatic executor of what the algorithm says.

Transparency and Accountability in Intelligent Systems

It is impossible to hold a system, or the institution using it, accountable if it is impossible to determine the data on which it relied, the purpose for which it was used, the extent of its accuracy, and who has authority to alter its decision.

For this reason, artificial intelligence governance has come to rely on a set of integrated elements, including:

Defining the legitimate purpose for using the system.
Assessing risks before deploying it.
Verifying the quality of the data used.
Documenting tests, results, and known technical limitations.
Identifying the person or entity responsible for operating the system.
Providing an appropriate level of human oversight.
Establishing procedures for addressing errors and complaints.

Accountability thus evolves from the simple question “Who made the mistake?” into a broader framework examining the design, training, operation, and oversight of the system.

Algorithmic Bias and the Principle of Equality

The fact that a decision is produced by an algorithm does not mean that it is neutral.

An algorithm learns from the data provided to it, and if historical data reflects patterns of discrimination or imbalance, the tool may reproduce those patterns more quickly and on a wider scale.

This problem is particularly evident in areas such as recruitment, credit scoring, insurance, and certain public services. Accordingly, evaluating a system should not be limited to its technical accuracy; its impact on different categories of users should also be measured, and the possibility of unlawful bias in its outcomes should be tested.

Algorithmic fairness is not achieved merely by treating all data in the same way. It also requires ensuring that the model does not produce unlawful discrimination as a result of its design or the data on which it was built.

Can Artificial Intelligence Be Granted Legal Personality?

In recent years, a strand of legal scholarship has questioned whether highly autonomous systems could be granted some form of legal personality, similar to the legal personality enjoyed by companies.

However, this approach raises fundamental issues. The legal personality of a company does not exist because it “thinks” like a human being, but because it is a legal instrument used to define rights, obligations, liability, and financial assets against which enforcement may be pursued.

Granting artificial intelligence an independent legal personality could, if the legal system were poorly designed, turn the technology into a barrier separating an injured party from the real persons responsible for developing, operating, or benefiting from it.

Accordingly, the more realistic approach remains the regulation of the liability of the humans, companies, and institutions that develop or use artificial intelligence systems, rather than granting the machine itself a legal status that enables liability to be shifted onto it.

Can Artificial Intelligence Become a Legislator?

Artificial intelligence can assist legislators with tasks that once required substantial time and effort, such as comparing thousands of legal provisions, identifying inconsistencies between laws and regulations, analyzing judicial decisions, studying legislative impact, comparing the experiences of different countries, and proposing alternative drafting.

However, drafting legislation is not merely a linguistic or statistical exercise.

Every piece of legislation involves choosing between conflicting interests, determining society’s priorities, and balancing freedom against security, worker protection against employer flexibility, privacy against the use of data, and innovation against the rights of creators.

These are political, legal, and ethical choices for which human beings must bear responsibility. Accordingly, the most logical model for the future is the legislator augmented by artificial intelligence, not the legislator replaced by artificial intelligence.

The UAE and the “Regulatory Intelligence” Experience

In 2025, the United Arab Emirates introduced an advanced experiment in this direction when the Cabinet approved the establishment of the Regulatory Intelligence Office and the launch of an integrated AI-powered regulatory intelligence ecosystem.

The concept is based on linking federal and local legislation with judicial decisions, procedures, services, and data, enabling the system to monitor the impact of laws, identify gaps and inconsistencies, and propose legislative updates.

The UAE Government announced that the use of these tools aims to accelerate the cycle of legislative research, drafting, evaluation, and issuance by up to 70%.

The significance of this experience lies in the fact that it does not grant the algorithm the authority to enact legislation, but rather uses it as a tool to improve Regulatory Intelligence, while constitutional legislative authority and decision-making remain with the competent human institutions.

Intellectual Property in the Age of Artificial Intelligence

Intellectual property is one of the fields in which artificial intelligence has raised some of the most difficult legal questions, whether in relation to content generated by models or works used to train them.

In January 2025, the U.S. Copyright Office issued Part Two of its report on artificial intelligence and copyright, concluding that using artificial intelligence as an assistive tool does not prevent protection, but material generated entirely automatically by the system without a sufficient degree of human creativity does not itself qualify for copyright protection under U.S. law.

The report also confirmed that merely writing instructions or Prompts is not, under current technological conditions, sufficient on its own to establish the required creative control, while elements created by a human, creative selection and arrangement, or substantial human modifications to outputs may qualify for protection.

This approach is more precise than treating all content involving artificial intelligence as either protected or unprotected in absolute terms, as the issue depends on the extent and nature of the human contribution.

Egyptian Law and Artificial Intelligence

As of August 2026, Egypt does not have a comprehensive horizontal law in force governing all applications of artificial intelligence comparable to the European Union’s AI Act.

However, this does not mean that there is a complete legal vacuum. Depending on their nature, uses of artificial intelligence are subject to a number of existing laws, most importantly:

Personal Data Protection Law No. 151 of 2020.
Anti-Cyber and Information Technology Crimes Law No. 175 of 2018.
Telecommunications Regulation Law No. 10 of 2003.
Intellectual Property Rights Protection Law No. 82 of 2002.
General rules governing civil liability, contracts, and consumer protection.
Sector-specific legislation in financial, banking, healthcare, and other fields depending on the nature of the application.

Accordingly, the absence of a “single artificial intelligence law” does not mean that the use of the technology falls outside the scope of legal responsibility.

Egypt Is Moving from Ethical Principles Toward a Governance Framework

Egypt began building its national artificial intelligence framework gradually. In 2023, the National Council for Artificial Intelligence issued the Egyptian Charter for Responsible AI, containing principles relating to fairness, transparency, accountability, privacy, security, and human responsibility.

The second edition of the National Artificial Intelligence Strategy 2025–2030 was subsequently issued, setting among its objectives the establishment of a national AI governance system, the development of a regulatory framework, and work toward drafting an Egyptian artificial intelligence law.

In March 2026, the Guide to Egypt’s National AI Governance Framework was issued. This represents an important development because it moves from general principles toward a more detailed methodology for classifying risks and establishing governance and compliance requirements.

The Guide is not, in itself, a binding law, but it expressly states that implementation of the governance mechanisms contained within it will be used in drafting Egypt’s future artificial intelligence law.

Egypt Is Moving Toward Risk-Based Regulation

Recent national documents reveal a clear direction toward adopting a risk-based regulatory philosophy, under which not all artificial intelligence systems are subject to the same regulatory burden.

This approach is appropriate to the nature of the market, as it allows innovation to be encouraged in low-risk applications while increasing obligations as the potential impact on safety, fundamental rights, or public interests rises.

In January 2026, the State’s institutional framework also evolved with the expansion of the Council’s mandate to become the National Council for Artificial Intelligence, Quantum Computing and Emerging Technologies, reflecting the widening scope of governance to include other highly advanced technologies.

What Should Egypt’s Artificial Intelligence Law Include?

The Office of Dr. Mostafa El Rouby – Attorneys and Legal Consultants believes that the anticipated Egyptian law should not merely be a copy of foreign legislation. Rather, it should draw on international experience while taking into account the particular nature of Egypt’s economy, administration, and society.

The most important issues that should be regulated include:

Defining the scope of the law and identifying the systems and entities subject to it.
Classifying systems according to risk instead of imposing the same obligations on all applications.
Defining legal liability among developers, suppliers, operators, and professional users.
Protecting data and privacy during the training and operational stages.
Addressing algorithmic bias and discrimination in decisions affecting individuals.
Imposing appropriate documentation and transparency requirements for high-risk systems without unjustifiably harming trade secrets.
Ensuring human oversight in sensitive decisions.
Regulating synthetic content and deepfakes in cases where they threaten rights or public trust.
Establishing intellectual property rules that take into account the use of artificial intelligence in creativity and training.
Creating regulatory sandboxes that allow companies to test new solutions under the supervision of competent authorities.

Transparency Does Not Mean Disclosing Trade Secrets

When drafting Egyptian legislation, it is important that the legislature does not overextend the concept of transparency.

Requiring every company to publish its algorithms or source code could threaten intellectual property, cybersecurity, and competitiveness.

A more balanced alternative is to grant regulatory authorities, where necessary, access to the technical information required for auditing, while requiring the system provider to give users understandable information regarding the nature of the decision, its principal limitations, and its risks, in proportion to the level of risk associated with the application.

Courts and the Legal Profession in the Face of Artificial Intelligence

The legal sector itself is one of the fields likely to undergo profound change as a result of artificial intelligence. Modern tools can assist with legal research, contract comparison, document review, detecting inconsistencies, and preparing initial drafts.

However, uncontrolled reliance on such tools creates significant risks, foremost among them reliance on incorrect information or judgments, disclosure of confidential client data, failure to understand legal context, and reducing professional judgment to an automated process.

Accordingly, lawyers must remain responsible for verifying every legal provision, judgment, or fact presented to a court or client, and an artificial intelligence tool must not become an independent source of legal trust.

Recommendations of the Office of Dr. Mostafa El Rouby

The Office of Dr. Mostafa El Rouby – Attorneys and Legal Consultants believes that the next phase calls for a number of practical priorities:

Completing the preparation of a national artificial intelligence law based on risk and consistent with the National Strategy and Egypt’s AI Governance Framework.
Defining a clear regulatory authority or regulatory framework with precise allocation of responsibilities between the central authority and sector-specific regulators.
Adopting the principle of human oversight for systems affecting rights, freedoms, or the fundamental interests of individuals.
Requiring entities to conduct impact assessments before deploying high-risk systems.
Developing liability rules so that an injured party knows which entity may be held accountable and responsibility is not lost between the developer, user, and system.
Training judges, lawyers, legislators, and experts to understand the nature of algorithmic systems and methods for evaluating evidence generated by them.
Reviewing the intellectual property and data framework to accommodate generative artificial intelligence without weakening creators’ rights or stifling innovation.
Supporting startups through regulatory sandboxes that allow safe innovation before imposing extensive compliance burdens.
Strengthening Arab and international cooperation because artificial intelligence systems and data do not stop at national borders.

Will Artificial Intelligence Write the Laws of the Future?

Artificial intelligence is likely to become increasingly present in the legislative drafting process. It will search, compare, analyze, monitor the impact of legal provisions, propose alternatives, and may help produce drafts that are more coherent and less contradictory.

But law is not merely a well-drafted text. It is an expression of society’s will, choices, and values, and it requires an authority that bears political and constitutional responsibility for those choices.

Accordingly, the issue is not whether artificial intelligence has the technical ability to propose a legal provision; it is already capable of contributing to that task. The real question is: Who decides that such a provision is fair and appropriate for society, and who bears responsibility for its consequences?

That responsibility must remain human.

Conclusion

The world has already entered a new era of “regulatory intelligence,” and it is now possible to use artificial intelligence to analyze legislation, assess its impact, and accelerate its preparation in ways that were previously unavailable.

However, the best legislation for the future will be neither law written by machines alone nor law that ignores the capabilities of technology, but rather law that combines the analytical power provided by machines with the legal and ethical judgment that remains a human responsibility.

For Egypt, the issuance of the National Artificial Intelligence Strategy 2025–2030, the Egyptian Charter for Responsible AI, and subsequently the National Governance Framework in 2026 demonstrates that the State has already moved from the stage of discussing the technology to the stage of building the rules governing it, while the issuance of comprehensive national legislation remains the anticipated step required to complete this framework.

The Office of Dr. Mostafa El Rouby – Attorneys and Legal Consultants – emphasizes that the challenge does not lie in resisting artificial intelligence or granting it unlimited authority, but in building a legal framework that enables innovation, defines responsibility, and protects human beings.

Artificial intelligence may help draft the laws of the future, but justice should not become an algorithm without accountability.