Introduction:
In the last two decades, the world has witnessed an accelerated transition from a traditional government model based on paper-based procedures and heavy administrative hierarchies to a digital government model that redesigns public service as a measurable and continuously improving user journey. This shift is no longer marginal; the average global e-government development index rose from 0.6102 in 2022 to 0.6382 in 2024, an increase that is associated with the expansion of online services, improved connectivity and the development of human capacity to support it.(United Nations, 2024). However, the digitization of services often means converting the transaction to a digital interface without changing the logic of the decision itself, while the smart state refers to a deeper level of aggregating data from multiple sectors and linking it to analytical and predictive algorithms to guide policies, and sometimes even to make procedural decisions automatically.
With the increasing use of algorithms in the management of education, health, justice, taxation, and security, questions of responsibility, justice, and privacy arise, especially when the tool shifts from improving service to redefining who holds decision-making power. Hence, the research question centers on whether countries are moving towards a fully automated administration or a hybrid model that redistributes roles between humans and machines within state institutions.
The paper starts from the thesis that the smart state is not just a technical modernization, but a political and ethical transformation that reshapes the state's relationship with the citizen by transferring part of the authority from the procedure to the algorithm, and from the transaction to the data file, which requires stricter governance rules, transparency and accountability.International indicators reveal that some Arab countries have become at the forefront of this process; Saudi Arabia ranked sixth globally in the EGDI 2024 with a score of 0.9602, while the UAE ranked eleventh with a score of 0.9533, reflecting that the region is not content with digitizing models but seeks large-scale data structures and operational platforms (UN eGovKB, 2024).
The first axis: From the bureaucratic state to the algorithmic state: the logic of transformation and its tools
The term smart state is used in contemporary literature to denote a state that bases its administration on data as a primary material for governance, and employs artificial intelligence and advanced analytics to increase the efficiency of services, guide policies and manage risks. This is different from e-government, which often focuses on digitizing service interfaces such as request, payment, booking and follow-up forms; e-government may move the form from window to site, but may keep the decision logic manual within the institution, while the smart state seeks to re-engineer the entire process through database integration, automation of approval and verification paths, and
Historically, the transformation can be read in three superimposed layers: first, paper-based management based on disparate files and long signature chains that raise the cost of the transaction and increase the likelihood of error and duplication; second, digitization that unifies service channels, enables electronic tracking and shortens completion time; and third, algorithmic enablement that transforms data from a subsequent record into instant fuel for decision-making, so that the state becomes able to monitor, measure and respond in near real-time.The Arab region shows clear examples of a shift in focus from simply providing a service to reducing paper and rebuilding processes: Dubai, as part of its Paperless Government strategy, has announced the reduction of more than 336 million sheets of paper and savings of more than 1.3 billion dirhams and more than 14 million dollars.(Dubai Media Office, 2021; Digital Dubai, n.d.). In Saudi Arabia, the size of the digital operation stands out as an indication that the platforms are no longer just communication channels but operating structures for the state; Absher platform transactions exceeded 430 million transactions in 2024 with hundreds of services, a metric that indicates that the state is building something like an operating system for citizens and residents (Saudipedia, 2024).
But the transition to a smart state requires a deeper technical structure than an application interface: there is a need for unified databases and trusted digital identity standards, operational linkage between multiple ministries and agencies to reduce redundancy and combat fraud, with national cloud architectures that guarantee data sovereignty and allow flexibility and expansion, in addition to a smart sensing and monitoring layer, cameras, sensors, Internet of Things, real-time records that feed predictive analytics.This is where big data plays the role of fuel; the greater the coverage, integration, and quality, the greater the state's ability to detect patterns: from traffic behavior and energy consumption to public health indicators and tax evasion.
The conclusion of this axis remains that automation appears to be an operational necessity in a fast and complex world, but it remains a political choice in terms of the limits of authorization for the algorithm, transparency and objection criteria, and the citizen's right to an understandable explanation of the decision.
Second Axis: The Smart State in Practice: Education, Health, Judiciary, and Taxation as Living Examples
If the infrastructure of the smart state is the nervous system, the education, health, judiciary, and tax sectors represent the testing grounds in which the difference between digitizing a procedure and automating a decision becomes clear. In education, the school transforms from a classroom in which the teacher writes down general grades to an accurate tracking system of the student's path through connected learning and evaluation platforms, so that educational data becomes a material for early diagnosis and personalization. The Saudi experience provides an example of the scale of this transformation; the digital bulletin issued by the digital sector regulator documents that the Madrasati platform recorded 7.2 billion visits, about 2.2 billion tests, 6.4 billion duties, and 537 million virtual classes
On a positive level, this volume allows building adaptive learning models that suggest individualized support paths, detect skill gaps early, and help decision makers direct resources, but it also opens a symbolic and practical risk of turning the student into a number on a dashboard, so that educational value is reduced to computerizable performance indicators, and educational pressure becomes related to what the algorithm measures rather than what is needed for human growth. In health, the smart state is manifested in the integrated digital health record and AI-powered diagnostic tools and resource management.The Emirates Health Services Corporation announced a plan to modernize radiology services that includes deploying AI solutions to detect breast cancer, pulmonary tuberculosis, stroke, osteoporosis, fractures, and chest diseases, with the aim of speeding up reporting and improving patient outcomes (Emirates Health Services, 2025).Here, the question of medical liability arises sharply: when the system suggests a reading of an X-ray image and then turns out to be wrong, who bears responsibility, the developer, the organization, or the doctor who approved the output? The question becomes even more sensitive with reports of high accuracy in some algorithms up to 99% in diagnosis, while emphasizing that the doctor's role in verification remains crucial (Khaleej Times, 2025).
In the judiciary, smart transformation takes two forms: first, making judicial services widely available digitally, reducing friction and enhancing procedural transparency; and second, introducing analytics to read patterns such as case distribution, adjudication times, and risk indicators. The Saudi Ministry of Justice reported that the Najiz platform provided more than 43 million services in the first half of 2024, a volume that reflects the transition from a paper-based institution to a connected services platform (Ministry of Justice, 2024).However, automated justice raises ethical issues when the algorithm shifts from organizing the file and prioritizing transactions to predicting the verdict or assessing the risk of litigants; even if the judge remains the final decision-maker, the model's recommendations may exert subtle pressure through so-called automation bias where humans tend to trust computer outputs.
In taxation, the path closest to sovereignty appears, because taxation affects the state's relationship with the economy and the citizen directly. Electronic invoicing, when linked to ZATCA's platforms through software interfaces and automated verification, opens the door to early detection of evasion and possibly dynamic tax pricing based on near real-time activity, as shown in the linkage and integration phases that aim to integrate invoicing solutions with ZATCA's systems (ZATCA, 2023).
In education, remote digital monitoring of exams, cheating detection analytics, and content recommendations, all of which theoretically increase integrity but expand the scope of surveillance into a space traditionally reserved for the learner. In health, AI is not limited to reading images, but extends to managing waiting lists, predicting demand for beds and medicines, and allocating staff, making first come, first served decisions vulnerable to numerical optimization if not controlled by clear fairness criteria.In the judiciary, projects to analyze crime patterns to direct patrols or identify focus areas are on the rise, an area where the risk of feeding growth with biased historical data and repeating discriminatory outcomes is high. In taxation, the integration of invoicing, customs data, and digital payments brings the state closer to a continuous accounting picture rather than an annual examination, a shift that may increase compliance and reduce the unregulated economy, but expands the state's ability to monitor the economy in real time.
Thus, it is clear that the smart state in application is not a single path, but rather a spectrum that starts with the digital service and ends with a fundamental question: do algorithms only manage procedures, or do they gradually begin to shape the content of the public decision through the ratings and expectations they produce?
The risks of the automated state and the future of governance: between efficiency and oversight
The first of these risks is privacy in the era of the smart state: as services become interconnected through a digital identity and unified records, the citizen tends to shift from being a requestor to being the subject of a data file that accumulates digital traces in education, health, taxation and transportation. This integration may reduce fraud and improve targeting, but it also raises the risks of excessive tracking, data leakage and overcoming the principle of proportionality, especially if the state moves from collecting data for service purposes to using it for other purposes without consent and clear criteria.
The second risk is algorithmic bias: models learn from the past, and if the past is characterized by discrimination or disparities in law enforcement or education and health opportunities, the algorithm may reproduce it in a seemingly neutral but unjust manner.Transparency and accountability are the cornerstone; who will be held accountable if the algorithm gets it wrong, or if an automated classification results in an individual being denied a service or placed under intense scrutiny? International reports indicate that building effective transparency tools in the public sector is a complex process that requires proactive tools such as public records of algorithmic systems, purpose and data labels, and rights impact assessments, not just publishing general principles (OECD, 2025).Practical government frameworks such as the Ethics, Transparency and Accountability Framework for Automated Decisions provide an example of how agencies can be required to define purpose, test fairness, document the decision, and provide understandable challenge channels for citizens (UK Government, 2023).
The more the state relies on standardized platforms and centralized algorithms, the more susceptible it is to becoming a total surveillance regime, where the technical ability to monitor and predict becomes a tempting alternative to political dialogue and addressing social causes. Cybersecurity risks are equally important. As the state moves towards interconnected digital platforms, software bugs or hacks can disrupt vital services or leak sensitive data on a large scale, making investment in protection and incident response a part of sovereign capacity rather than a technical item.The risks of relying on specific technology providers through long contracts and closed standards, which may restrict the independence of public decision-making and make it difficult to audit models and data sources. From a social justice perspective, the smart state may deepen the digital divide if it is not accompanied by digital inclusion policies, as those who do not have an effective digital identity, stable connectivity or usage skills will find themselves out of service or under a non-compliant classification.Therefore, the algorithmic accountability literature proposes a complementary set of mechanisms: independent audits, disclosure of purposes and data, rights impact assessments, grievance pathways that enable the individual to understand and challenge the decision, and public monitoring tools that foster trust (Open Government Partnership, 2021).
In terms of scenarios, three paths can be envisioned: a fully automated state that leaves systems with ample space to determine eligibility and risk; a hybrid state that integrates AI as a supportive decision-making tool while keeping humans responsible and accountable for the final decision; and a third path led by social and judicial resistance that restores human decision-making by imposing limits on algorithmic authorization and raising transparency and objection requirements.
Conclusion
This paper concludes that the smart state is not just a later stage of e-government, but a reformulation of the governance equation by transferring part of the authority from the human procedure to the algorithmic system. When databases are integrated and digital flows are organized in education, health, justice and taxation, services are accelerated and the ability to measure and plan improves, as shown by Arab experiences in reducing paper, expanding platforms and raising digital readiness indicators.However, this gain does not come without a price: data integration expands the possibilities of monitoring and makes privacy an issue of judgment, algorithmic bias may turn social discrimination into a mathematical result, and lack of transparency may create an effective but unaccountable state.
Therefore, the answer to the research question tends to suggest that the nearest future is not a fully automated administration, but a hybrid model, in which algorithms take over the work of screening, verification, early detection, and optimizing the flow of resources, while decisions affecting rights, entitlement, and punishment remain subject to clear human responsibility that can be scrutinized and challenged.In other words, what is needed is not a state without humans, but a state that uses AI without compromising human responsibility: through rigorous data governance, rights impact assessments, public records of automated systems, an understandable right of interpretation for citizens, and effective channels of challenge. International governance experience shows that tools like algorithmic transparency, proactive disclosure, model documentation and auditing are not a theoretical luxury, but a requirement to build public trust and prevent efficiency from becoming censorship (OECD, 2025; UK Government, 2023).
In a smart state, who writes the rules that actually govern us, the parliament that legislates and defines boundaries, or the algorithm that translates the rules into classifications and predictions that may gradually become the invisible law?
Comments