Artificial Intelligence Platforms in Education

Kosorukov Artem Andreevich ORCID: 0000-0002-0275-4899 PhD in Politics Associate Professor; Department of Political Analysis; Lomonosov Moscow State University 119991, Russia, Moscow, Lomonosovsky Prospekt, 27/4, office G645 kosorukov@spa.msu.ru

Introduction

Intelligent technologies in education play a key role in the development of modern society, forming the basis for sustainable economic growth and social stability. Historically, the introduction of new technologies into the educational process has contributed to increased access to knowledge, increased literacy, and the creation of a more skilled workforce. For example, the invention of the printing press in the 15th century became a catalyst for the spread of knowledge, which led to the age of Enlightenment and accelerated scientific and technological progress. In the 20th century, the massive introduction of computers and the Internet into educational institutions opened up new horizons for distance learning and global knowledge exchange.

In today's world, artificial intelligence (AI) technologies and digital platforms are transforming the educational process, making it more personalized, accessible, and effective. This is especially important in the context of globalization and the digital economy, where the competitiveness of countries is determined by the level of training of their citizens. For example, adaptive educational systems such as Knewton and DreamBox Learning take into account the individual needs of students, which contributes to deeper learning and the development of critical thinking. Such approaches not only improve the quality of education, but also promote social mobility by providing equal learning opportunities regardless of geographical location or social status.

The economic importance of technology in education lies in the training of qualified personnel for high-tech industries. Modern educational platforms such as Knewton and DreamBox Learning use AI to analyze data on student progress, which allows them to optimize curricula and increase their effectiveness. This, in turn, contributes to the growth of labor productivity and the development of an innovative economy. From a political point of view, technology in education contributes to the development of critical thinking, which is necessary for active and competent participation in social processes.

The degree of scientific elaboration of the problem

Russian research in the field of artificial intelligence in education focuses on studying its integration into the educational process, analyzing prospects and identifying possible risks. For example, S.M. Kashchuk's article "Artificial Intelligence in education: what to fear, what to use" ^[1] emphasizes the need for a cautious approach to the implementation of AI, focusing on ethical aspects and possible errors of algorithms. Bukina T.V. in her work "Artificial Intelligence in Education: current state and development prospects" ^[2] analyzes key areas of AI application, including adaptive learning and automated assessment. Yartseva E.Ya.'s research "Integration of artificial intelligence into education" ^[3] highlights the importance of creating hybrid models combining AI and traditional teaching methods. The article by E.A. Pospelova et al. "Generative artificial intelligence in education: analysis of trends and prospects" ^[4] discusses the possibilities of using generative models to create educational materials. Kalinin A.A. and others emphasize the advantages of preparing educational content using domestic tools and technologies of generative AI ^[5]. These works highlight the importance of AI for personalizing learning, but also point to the need to develop standards and regulatory mechanisms.

Foreign studies focus on global trends and innovative approaches to the use of AI in education. For example, in the article "Generative AI Tools in Education: A Survey of Educators' Attitudes" ^[6], Ashish D. examines teachers' perceptions of generative AI tools, identifying their potential to increase student engagement. Tsiomin Ts. The work "Large-Scale Assessment in Science Education" ^[7] analyzes the use of AI for large-scale assessment of knowledge, emphasizing its role in the standardization of educational processes. Omodan B. and Marongwe N. in the article "The Role of Artificial Intelligence in Decolonizing Academic Writing" ^[8] consider AI as a tool for creating inclusive educational content. Eslit E. in her work "Teachers' Reflections on Literature and Language Education in the Era of Artificial Intelligence" ^[9] emphasizes the importance of preserving the humanitarian approach in the context of digitalization. Ciah K. reveals the specifics and risks of implementing AI in secondary schools ^[10], demonstrating that AI not only improves the quality of education, but also helps to address global challenges such as accessibility and inclusivity of educational products.

Research methodology

To conduct an expert survey of 16 leading Russian experts representing higher education institutions from 8 federal districts of the Russian Federation, 7 equilibrium analysis criteria were developed that made it possible to evaluate the effectiveness and functionality of the AI platforms listed above in education. These criteria include:

1. Adaptability – the ability of the platform to adapt to the individual needs and level of knowledge of students,

2. Interactivity – the level of interaction between the platform and the user, including feedback and engagement,

3. Functionality – the main features of the platform and the tasks it solves,

4. Effectiveness – the impact of the platform on learning outcomes and student achievement,

5. Availability – price, technical requirements and ease of use,

6. Integration – the ability to integrate with other educational systems and platforms,

7. Innovation – the use of advanced technologies and unique functions.

Based on the criteria outlined, each of which was evaluated on a scale of "low -moderate - medium – high", experts evaluated the following platforms: Knewton, DreamBox Learning, Civitas Learning, IBM Watson Education, ProctorU, ExamSoft, Turnitin, Grammarly, Antiplagiat.ru , Edsight, Automated Essay Scoring (AES). The results of summarizing the expert data are shown in table 1 (the weighting of the criterion on the scale "low -moderate - medium – high" was carried out according to the principle of relative majority, for example, if 7 experts voted for the indicator "high", and fewer votes were cast for each of the other indicators, then the indicator "high" was indicated in the table).

The purpose and objectives of the study

The purpose of the study is to conduct a comprehensive analysis of the use of artificial intelligence platforms in the field of education to assess their impact on the quality of educational services, the effectiveness of educational processes, as well as to identify the prospects and ethical aspects of their integration.

To achieve this goal, the following research objectives are set:

1. To study the historical stages of the development of AI technologies in education and their role in the digital transformation of educational processes,

2. To conduct a comparative analysis of modern AI platforms in the field of adaptive learning, proctoring, text verification and evaluation of creative works,

3. To identify the advantages and limitations of using AI platforms, including the problems of adapting to the individual needs of students, ethical risks and difficulties in evaluating creative work,

4. Evaluate the effectiveness of the use of AI platforms in education based on expert survey data and comparative analysis,

5. To identify the prospects for further research in the field of AI adaptation to the emotional and cognitive characteristics of students and integration with IoT and blockchain technologies.

Educational platforms using artificial intelligence technologies

One of the first educational computer systems created in the 1960s at the University of Illinois was PLATO (Programmed Logic for Automatic Teaching Operations). PLATO was developed in 1960 under the supervision of Professor Donald Bennett and his team at the University of Illinois. Initially, it was aimed at automating the educational process and providing students with the opportunity to interact with educational materials. The system has gone through many versions and updates, starting with the simplest text interfaces and ending with more complex graphical interfaces in the 1980s. PLATO provided students with the opportunity to interact with training courses, take tests, complete assignments, and receive real-time feedback. She also supported student-to-student communication functions, including chat rooms and forums, which facilitated knowledge sharing and collaborative learning. PLATO was one of the first systems to use graphics and animation to create a more engaging and effective educational process. PLATO has inspired the creation of educational systems such as MOOCs (massive open online courses). Despite the fact that the system itself ceased to exist in 2006, its ideas continue to live on in modern educational platforms and tools that continue to evolve and adapt to the needs of students and teachers.

In the 1970s, the SHRDLU project was developed, which became one of the first systems using artificial intelligence elements for natural language processing. This project, developed at the Massachusetts Institute of Technology, has become a significant step forward in the field of technologies that allow computers to interact with people in their native language. SHRDLU had the ability to interpret and process commands formulated in English, which provided users with the ability to interact with the system. The project was carried out in a limited environment — a simple three-dimensional space consisting of blocks that users could move, combine and manipulate, which allowed the system to interpret commands and perform appropriate actions. SHRDLU could conduct dialogues with users, answering their questions and clarifying the commands received, which created the impression of communicating with an intelligent computer. The system had the ability to remember the specifics of a conversation with a specific user, which allowed it to support more complex interactions and better understand requests.

In the late 1980s and early 1990s, intelligent learning systems, known as ITS (Intelligent Tutoring Systems), began to actively develop, which were able to adapt to the students' level of knowledge and provide them with individually tailored assignments. One of the most notable examples of ITS is the Cognitive Tutor system, developed in the 1990s at Carnegie Mellon University. This system had the ability to analyze the student's level of knowledge and skills, adapting learning materials and assignments to individual needs, which allowed students to study at their own pace and focus on areas where they had difficulties. Cognitive Tutor offered a variety of tasks, including simple exercises to consolidate basic concepts, as well as more complex tasks aimed at developing critical thinking and problem-solving skills. The system provided instant feedback on completed tasks, which helped to correct mistakes and deepen understanding of the material, as well as help students realize their strengths and weaknesses. Cognitive Tutor used knowledge models that described exactly what the student knows and what they still have to learn. This system also included elements of interactivity, allowing students to actively participate in the learning process through simulations, games, and other interactive components. Cognitive Tutor applied cognitive models that helped to understand how students solve problems and what strategies they use. At the same time, teachers received reports on the progress of their students, which allowed them to better understand in which areas students were experiencing difficulties and how the learning process could be improved.

As a result, by the beginning of the 21st century, artificial intelligence technologies were being actively introduced into the field of education, and the use of educational AI technologies on digital platforms covered a variety of areas, such as adaptive learning, proctoring, text verification, and creative work evaluation.

1. Adaptive learning is a technique based on the use of artificial intelligence, which adjusts the educational process taking into account the individual needs and level of knowledge of each student.

Knewton is a platform focused on adaptive learning that uses modern data analysis and artificial intelligence technologies to create a personalized educational experience for students. Launched in 2008, it has quickly gained a reputation as one of the leading players in the field of adaptive educational technologies. The platform collects and processes data on student achievements in real time, including information about completing assignments, challenging topics, and interacting with learning materials. Based on this information, Knewton adjusts study materials and assignments to the individual needs of each student. If the system detects that a student is having difficulties in a particular area, it can offer additional resources such as videos, articles, or interactive assignments. After analyzing the learning progress, Knewton provides recommendations on the materials studied. The platform can be integrated with various educational systems and learning management platforms, which allows educational institutions to use its capabilities within existing courses and programs. Teachers have access to data on student progress, which helps them better understand which areas students are experiencing difficulties in and how the quality of the educational process can be improved. For example, if a student is not good at algebra tasks, Knewton can offer additional exercises on the basics of this discipline, as well as videos explaining theories and methods for solving various types of problems. If a student has difficulty understanding plant and animal life processes in a biology course, the system can provide additional resources such as interactive diagrams or animations to help visualize the material being studied and explain complex aspects of the topic.

DreamBox Learning is an innovative educational platform specifically designed for learning mathematics among elementary and secondary school students. Using advanced artificial intelligence and adaptive learning technologies, this platform forms individualized learning paths, which allows each student to master the material at a pace convenient for him and in accordance with his personal needs. DreamBox monitors students' actions in real time, including their responses to assignments, time spent completing tasks, and applied solution methods. Based on the information collected, the system adjusts the content and difficulty level of assignments to match the individual abilities of each student. The platform develops unique learning paths for each student, taking into account their strengths and weaknesses. This allows students to focus on those aspects that require additional attention and progress through the program. DreamBox offers a wide range of interactive tasks and games, which makes the learning process more interesting and exciting. The platform also provides teachers and parents with detailed reports on student progress, which allows them to track their achievements, identify areas that require additional work, and adjust curricula if necessary. For example, if a student has difficulty understanding fractions, DreamBox offers additional tasks using visual models and interactive tools for deeper learning. The platform includes game elements, allowing students to earn points and achievements for successfully completing tasks, which motivates them to continue learning. DreamBox Learning is actively used both in educational institutions and at home to improve the level of mathematical literacy. The platform promotes the development of critical thinking and problem-solving skills among students, as well as provides the necessary support in studying mathematics, which in turn improves academic performance and increases self-confidence ^[11].

Civitas Learning is a platform that uses analytical tools based on artificial intelligence to predict student learning outcomes and identify those who are experiencing difficulties in the learning process. The main objective of this platform is to help educational institutions more effectively support their students and increase their chances of successful completion of the curriculum. The platform has a multi-level structure for data collection and processing, aggregating a variety of educational data from various sources. These include academic indicators (academic performance, intermediate assessment results, activity in learning management systems), behavioral aspects (attendance, participation in discussions, interaction with digital resources) and contextual factors (socio-demographic characteristics, information about extracurricular activities and psychometric data). To develop predictive models, Civitas Learning uses hybrid machine learning algorithms such as gradient boosting (for example, XGBoost and LightGBM) to analyze nonlinear dependencies, deep neural networks (for example, LSTM and Transformer) to work with time series of academic activity, as well as ensemble methods (such as Random Forest and Stacking) to reduce generalization errors. The platform collects and analyzes data about students, including their academic performance, class attendance, participation in extracurricular activities, and other factors that affect learning outcomes. Civitas Learning uses machine learning algorithms to create predictive models that help educational institutions identify students in need of additional support. Civitas Learning models take into account many variables, which makes it possible to more accurately predict risks and intervene at the early stages of their manifestation, offering assistance to those students who are at risk. This may include additional counseling, mentoring, access to learning resources, or other forms of support. The platform also provides personalized recommendations for students based on their unique needs and profiles, which helps them better navigate the learning process and find the necessary resources for successful learning. The platform provides educational institutions with resources for generating reports and visualizing information, which enables administrative staff and teachers to better analyze changes in student performance and make more informed decisions.

IBM Watson Education is a platform based on artificial intelligence technologies designed to support the educational process. It offers a variety of tools and functions aimed at improving the quality of learning and optimizing educational activities. Using machine learning algorithms, the platform analyzes data about students and their academic performance, providing personalized recommendations on the use of additional educational resources. This allows students to have instant access to materials that match their level of knowledge and learning style. The system collects and processes information about students' progress, including test scores and their participation in classes. The platform includes interactive tools that promote greater student engagement in the learning process, such as mobile apps, business games, and business simulations, making learning more interesting and engaging. Watson Education provides teachers with opportunities to develop and manage curricula, as well as monitor student progress, which helps them organize classes more effectively and approach each student individually. The platform offers access to an extensive set of educational materials, including articles, audio and video content, which allows students to deepen their knowledge and broaden their horizons. IBM Watson Education provides personalized recommendations and data analysis, which contributes to more effective work for both teachers and students. By analyzing data on student progress, teachers can quickly adjust their teaching methods. Automation of processes and the selection of additional materials allow them to focus on more significant aspects of education, such as interaction with students and the development of new teaching approaches.

2. Proctoring is a system of control over the conduct of exams aimed at eliminating cases of cheating. The use of artificial intelligence technologies makes it possible to automate this process, which contributes to maintaining safety and fairness in the assessment.

ProctorU is a remote proctoring platform that uses advanced artificial intelligence technologies to ensure the safety and integrity of exams conducted online. It is aimed at educational institutions and organizations that strive to ensure that students pass exams without any violations. The platform monitors students during the exam using webcams and microphones, which allows proctors to monitor student behavior in real time and identify suspicious activity. Using artificial intelligence algorithms, ProctorU analyzes video and audio data, which makes it possible to identify unusual movements, sounds, or interactions that may indicate the use of prohibited materials or assistance from third parties. All examination sessions are recorded, which provides an opportunity for subsequent analysis and verification for violations, as well as allows you to re-view the videos in case of questions about the integrity of the exam. The platform offers a user-friendly interface for both students and teachers. Students can easily register for exams, and teachers have the ability to configure proctoring settings and receive reports on student behavior and academic performance. ProctorU supports a variety of exam formats, including multiple choice tests, as well as essays and practical assignments. ProctorU-like services play an important role in ensuring integrity in online education, helping to minimize the risks of fraud and create a level playing field for all participants, which, in turn, helps to increase trust in online exams in general ^[12].

ExamSoft is a system designed for the organization of exams, which offers proctoring solutions using artificial intelligence technologies. The platform is aimed at educational institutions and professional organizations that strive to ensure the integrity and safety of the exam process. ExamSoft has the ability to monitor student actions on the screen during the exam, including recording attempts to switch to other applications or web resources. The platform is also capable of recording video and audio during the exam, enabling proctors and administrators to analyze student behavior and identify possible violations. In addition, ExamSoft offers tools for analyzing exam results: teachers can receive detailed reports on students' progress and achievements, as well as identify weaknesses in their knowledge and adjust curricula based on the data obtained. The platform provides secure access to exams, preventing unauthorized access, and uses various authentication and information security methods. ExamSoft can integrate with other educational systems and learning management platforms, which simplifies the exam administration process. A number of educational institutions, including medical and law universities, use ExamSoft to conduct certification exams, where high standards of honesty and security are especially important. Higher education institutions use ExamSoft to monitor online exams, which allows them to maintain academic integrity and ensure a high level of education.

3. The use of artificial intelligence for text analysis provides an opportunity for teachers and students to evaluate the quality of created materials, detect plagiarism and receive advice on their improvement.

Turnitin is a platform that uses artificial intelligence to analyze texts for plagiarism and evaluate their uniqueness. It is widely used in educational institutions, contributing to maintaining academic integrity and improving the quality of students' written work. The system compares student papers with an extensive database that includes millions of educational materials, articles, online resources, and previously submitted texts. After the verification is completed, Turnitin generates a report that indicates all the matches with the sources, while it includes the percentage of originality of the text and links to original materials, which allows both teachers and students to easily identify problematic issues. In addition, Turnitin is not limited to detecting plagiarism, it also offers recommendations for improving the style and structure of the text, including tips on paraphrasing, strengthening arguments and improving readability. The platform can be integrated with various learning management systems such as Moodle, Blackboard and Canvas, which makes the process of uploading work and receiving reports more convenient. Turnitin offers feedback tools that allow teachers to leave comments and evaluate students' work, and supports a variety of file formats, including text documents, PDFs, and presentations, which ensures its versatility in various educational contexts. Many educational institutions use Turnitin to check term papers and theses, which helps maintain high standards of academic integrity. This system plays a significant role in fostering a culture of responsibility and integrity among students. The use of Turnitin helps not only in detecting plagiarism, but also in developing critical thinking and writing skills, which is an important aspect of the educational process.

Grammarly is a highly effective tool that uses artificial intelligence technologies to analyze grammar, style, and clarity of texts. This platform is designed to help users create better and more professional written works. Grammarly automatically detects grammatical errors, typos, and punctuation issues, offering corrections and explanations, which helps users learn based on their mistakes. The tool also analyzes writing style, providing recommendations for improving wording, word choice, and sentence structure. Thanks to this, Grammarly helps to make the text more understandable and logical by offering simplifications and clarifications. This tool adapts to users' individual writing styles and provides personalized advice, allowing everyone to develop their writing skills according to their unique needs. Grammarly is available as a web application, as well as browser extensions and applications for Microsoft Word and Google Docs, making it convenient to use in various software environments. The paid version of Grammarly has a plagiarism check feature that compares text with millions of web pages and other sources, helping users verify the originality of their work. The platform provides users with feedback on their progress and offers learning resources that promote writing skills and improve general literacy. Grammarly helps students improve their essays, term papers, and other written assignments, while many professionals use it to prepare business letters, reports, and presentations to create clearer and more professional documents. Writers and content creators use Grammarly to edit their texts, improve readability, and improve the quality of the content they create. Using this tool helps users not only to correct mistakes, but also to realize how they can improve their texts in the future.

Antiplagiat.ru It is a platform that uses advanced technologies, including artificial intelligence, to analyze texts for plagiarism and determine their uniqueness. The platform is widely used in educational institutions, scientific institutions and among content authors, contributing to ensuring the originality and high quality of the created materials. Antiplagiat.ru Analyzes texts for borrowings from a variety of sources, such as web pages, scientific articles, and other publications. The platform provides users with a percentage of the text's originality, which allows them to assess how unique their work is. This is especially true for students and teachers who are required to adhere to the principles of academic integrity. Antiplagiat.ru It uses algorithms that are able to detect not only direct borrowings, but also paraphrased phrases, which makes the verification process more detailed and reliable. The platform has an intuitive interface that allows users to easily upload their texts and receive analysis results. After the verification is completed, users receive comprehensive reports indicating all detected borrowings with links to sources, which makes it possible to make the necessary changes and enhance the originality of their work. Antiplagiat.ru It is actively used in educational institutions to check term papers and theses, and also supports various document formats such as Word, PDF and text files, which makes it a universal tool for various situations. There are many ways to apply Antiplagiat.ru : Students check their term papers, theses, and tests for uniqueness before submitting them to an educational institution, teachers use the system to check students' work and ensure that standards of academic integrity are met, and online content authors check articles, blogs, and other materials for originality before they are published.

4. Evaluating creative works, including essays and art projects, is a time-consuming task. However, with the help of artificial intelligence technologies, it is possible to automate this process and provide more objective results.

Edsight is an educational platform that uses artificial intelligence technologies to analyze written works. It is aimed at both students and teachers, offering a detailed analysis of texts based on predefined criteria. Using AI algorithms, this platform performs automated evaluation of written works, which allows for fast and efficient analysis of various parameters, including structure, grammar, style, and content. Edsight provides constructive feedback for both students and faculty. Students can get tips on improving their texts, and teachers can get information about the quality of assignments in the classroom. The platform has the ability to adapt to different levels of student training and the specific requirements of educational programs, which makes it a universal tool for educational institutions. Edsight evaluates texts based on pre-established criteria such as presentation logic, argumentation, use of sources, and adherence to format, which helps ensure objectivity and consistency in the evaluation process. The platform allows you to download and analyze works in various formats, including Word documents, PDFs, and text files, and offers an intuitive interface that makes it easier to download materials and obtain analysis results without the need for complex settings. In addition, Edsight can be integrated with other educational systems and tools, which simplifies the assessment and management of the learning process. Students use the platform to self-check their written papers before submitting them, receiving recommendations for improvement. Teachers use Edsight to evaluate students' work, which allows them to reduce the time for verification and focus on more important aspects of learning. Educational institutions are implementing this platform in the educational process to improve the quality of learning and ensure the objectivity of assessment.

Automated Essay Scoring (AES) is a system that applies algorithmic methods to evaluate written papers such as essays. This system analyzes texts, taking into account many aspects, including their structure, grammatical features, style and content. AES uses advanced algorithms and machine learning models for word processing, based on pre-established criteria such as logical organization of thoughts, argumentation, language design and grammatical constructions. The essay is evaluated based on various parameters, including structure, grammar and spelling, as well as content (relevance to a given topic, depth of analysis, use of examples and proofs), style (variety of vocabulary and complexity of syntax) and other elements. Despite the fact that AES is not able to completely replace the assessment conducted by a teacher, it can be a valuable tool for preliminary analysis of work. This system offers students feedback, highlighting both the strengths and weaknesses of their essays. The use of AES significantly reduces the time required to evaluate written papers, which is especially important in large study groups or during mass testing. However, AES may not always correctly interpret the context or tone of the text, which sometimes leads to errors in evaluation. In particular, the algorithms of the system practically do not take into account creative approaches or non-standard solutions presented in the works. For a final assessment, especially in cases requiring in-depth analysis and understanding, it is still necessary to involve an expert. AES finds application in various standardized tests, such as TOEFL and GRE, to evaluate participants' written papers. Moreover, many educational institutions are introducing AES into their educational programs in order to improve the learning process and evaluate students' writing skills.

The results of the study

After analyzing the data from an expert survey conducted taking into account the assessment of each of the criteria on a scale of "low -moderate - medium – high", the author compiled table 1.

Table 1. Results of the expert survey.

The analysis of the results of educational AI platforms, according to the results of the study, showed the following results for each of the criteria:

1. Adaptability: the most adaptive platforms are recognized Knewton, DreamBox Learning, and IBM Watson Education. They actively use AI to personalize learning. Platforms such as ProctorU and Turnitin have low adaptability, as their functions are limited to monitoring and verification.

2. Interactivity: DreamBox Learning is highly interactive due to game elements and visualization. Platforms for text verification such as Turnitin and Antiplagiat.ru , have a low level of interactivity.

3. Functionality: all platforms have received high marks for functionality, as each of them solves specific tasks in education.

4. Efficiency: All platforms have shown high efficiency in their fields, whether it's adaptive learning, proctoring, or text verification.

5. Accessibility: Grammarly received the highest score for accessibility due to its free version and simple interface. IBM Watson Education and Civitas Learning have higher technical requirements and cost, which reduces their availability.

6. Integration: All platforms support integration with other systems, making them universal tools for educational institutions.

7. Innovation: Knewton, DreamBox Learning, and IBM Watson Education have received high marks for their use of cutting-edge technology. Platforms for proctoring and text verification, such as ProctorU and Turnitin, are less innovative, as their functions remain unchanged over time.

Based on the cumulative assessments received from experts for each of the platforms, an integrated multi-criteria assessment was carried out, which allowed us to build a platform rating. To build a rating, each platform is assigned points on a scale: high - 4 points, medium - 3 points, moderate - 2 points, low - 1 point. The sum of the points according to all criteria (adaptability, interactivity, functionality, efficiency, accessibility, integration, innovation) determines the platform's place in the rating.

Platform rating:

1. DreamBox Learning. Total points: 27. Grades: high (6 criteria), average (1 criterion). The leader of the rating due to its high interactivity and innovation.

2. Knewton. Total points: 25. Grades: high (5 criteria), moderate (1 criterion), average (1 criterion). Strong adaptability and functionality, but moderate interactivity.

3. IBM Watson Education. Total points: 25. Grades: high (5 criteria), moderate (1 criterion), low (1 criterion). High marks for innovation, but low availability.

4. Grammarly. Total points: 24. Grades: high (4 criteria), moderate (2 criteria), average (1 criterion). Better accessibility, but moderate adaptability and interactivity.

5. Automated Essay Assessment (AES). Total points: 21. Grades: high (3 criteria), moderate (2 criteria), low (2 criteria). High functionality, but low interactivity.

6. Civitas Learning. Total points: 20. Grades: high (3 criteria), moderate (2 criteria), low (2 criteria). Efficiency and integration are good, but there is little interactivity.

7. ExamSoft. Total points: 19. Grades: high (3 criteria), moderate (2 criteria), low (2 criteria). It is well suited for proctoring, but it is not interactive enough.

8. Edsight. Total points: 19. Grades: high (3 criteria), moderate (1 criterion), low (3 criteria). A narrow focus on evaluating texts.

9. ProctorU. Total points: 18. Grades: high (3 criteria), medium (1 criterion), low (3 criteria). Low adaptability and interactivity.

10. Turnitin. Total points: 18. Grades: high (3 criteria), medium (1 criterion), low (3 criteria). It is effective in checking for plagiarism, but it does not adapt to students.

11. Antiplagiat.ru Total points: 17. Grades: high (3 criteria), medium (1 criterion), low (3 criteria). Similar to Turnitin, but with less innovation.

Thus, an expert survey showed that DreamBox Learning, Knewton, IBM Watson Education combine high adaptability, innovation and functionality, while platforms for proctoring (ProctorU, ExamSoft) and text verification (Turnitin, Antiplagiat.ru They lag behind in adaptability and interactivity. Grammarly stands out for its accessibility and the balance between functionality and ease of use.

Discussion of the results

The conducted research made it possible to identify key trends and contradictions related to the introduction of AI platforms into the educational process. The results of the expert survey showed that adaptive learning-oriented systems (DreamBox Learning, Knewton, IBM Watson Education) received the highest marks due to their ability to personalize learning trajectories and use advanced data analysis algorithms. However, platforms that perform highly specialized tasks (proctoring, text verification), such as ProctorU and Turnitin, have demonstrated limited adaptability and interactivity, which reduces their potential for integration into flexible educational models.

An interesting aspect was the contradiction between the functionality and accessibility of the platforms. For example, Grammarly, which received the highest score in accessibility, is inferior to the leaders in adaptability, which emphasizes the need for a balance between ease of use and complexity of technological solutions. In addition, the revealed low innovativeness of text verification platforms indicates conservatism in the development of these tools, despite their demand.

Experts also noted the ethical risks associated with algorithm bias and automation of educational processes. For example, essay assessment systems (AES) demonstrate high performance in standardized tests, but are unable to account for creativity and context, which calls into question their universality. These challenges require the development of hybrid models that combine AI with human control.

The results of the study are consistent with global trends described in foreign works, such as the need for inclusivity and overcoming digital inequality. However, the Russian context highlights the importance of adapting international solutions to local educational standards and infrastructural constraints.

Conclusions

The technological leadership of adaptive platforms, in particular, adaptive learning platforms (DreamBox Learning, Knewton), is confirmed in the study by the fact that they demonstrate their effectiveness through the personalization of the learning process, real-time data analysis and integration with existing educational systems. The success of these platforms in the expert survey is due to their high innovation and functionality. The limitations of specialized tools, in particular, proctoring and text verification systems (ProctorU, Turnitin) remain narrowly focused, which reduces their role in the integrated digitalization of education, and their further development requires the introduction of adaptability and increased interactivity. The ethical and methodological challenges in the process of introducing educational platforms based on artificial intelligence technologies are that the widespread adoption of AI is faced with problems of algorithm bias, lack of interpretability of solutions, and risks of dehumanization of education. In this regard, the development of hybrid models combining AI with expert assessment is a key methodological solution that reduces ethical problems. The prospects for integrating platform technologies in education look quite high, so further research should focus on integrating AI with IoT and blockchain to improve data security, as well as on creating algorithms that take into account the emotional and cognitive characteristics of students. Educational institutions should combine the use of educational AI platforms with maintaining the role of the teacher as a moderator of the educational process, ensuring a balance between automation and human interaction.

Conclusion

The conducted research has confirmed that artificial intelligence technologies have a significant impact on the development of education, contributing to its digital transformation. In the educational field, intelligent learning systems such as Knewton, DreamBox Learning, and Squirrel AI, as well as automated assessment platforms including Turnitin, Grammarly, and Automated Essay Scoring (AES), provide personalized learning, learning process adaptation, and objective assessment of knowledge. Nevertheless, the large-scale implementation of artificial intelligence in education faces a number of serious challenges. One of the key difficulties is the lack of adaptability of AI systems to the individual cognitive characteristics of students. Modern learning platforms are often limited in taking into account students' motivation, emotional state, and style preferences, which can negatively affect their effectiveness. In addition, automated knowledge assessment systems, despite their high level of accuracy, face challenges when reviewing creative works, complex essays, and interdisciplinary projects where subjective factors play an important role. The ethical aspects of using AI are also of considerable concern. Algorithms based on training data can inherit biases, which leads to an uneven distribution of educational opportunities. For example, algorithms trained on limited or biased data may underestimate the grades of students from different socio-economic groups or cultural contexts. In addition, there is a risk of excessive automation of the educational process, in which the role of the teacher is reduced to a moderator, and live interaction loses its importance.

Taking into account current trends, artificial intelligence will continue to transform education, creating new opportunities for personalization, automation and improving the quality of services provided. However, the successful implementation of these technologies requires further research aimed at overcoming existing limitations and ensuring a balanced interaction between artificial intelligence and humans, including:

1. Development of hybrid AI models combining machine learning and expert systems to increase transparency and explainability of decisions,

2. Research of AI adaptation methods to the individual needs of users, analysis of the emotional and cognitive state of students,

3. Improving data processing algorithms in order to minimize bias and ensure their accuracy and interpretability,

4. Development of standards and regulatory mechanisms for the safe and ethical implementation of AI in socially significant areas, contributing to increased trust in technology,

5. Integration of AI with the Internet of Things (IoT) and blockchain to increase the effectiveness of monitoring and data protection, opening up new horizons for the safe use of AI technologies.