Artificial intelligence in sociology and social policy Open Access

Artificial intelligence is a topic of immense interest due to recent rapid technological advances including cloud and quantum computing (Badawy et al., 2025). Knani et al. (2022, p. 1) state that “AI comprises other intelligent technologies and tools, machine learning, the Internet of Things, artificial neural networks, big data, smart robots, and virtual and augmented reality applications”. Intelligence in general refers to how a living being deciphers information in order to obtain knowledge. Traditionally, intelligence has been considered in terms of language and mathematics, but more recently emotional forms have been incorporated into revised definitions. The emphasis on any form of intelligence is to be able to solve problems and make correct decisions. In artificial forms of intelligence, this means using some form of technology such as computers and machines to solve things. This implies that human learning and responses can be simulated through technology.

Artificial intelligence is defined as using machines to simulate human intelligence processes that include problem solving, drawing inferences and providing feedback (Banna and Alam, 2025). Overall societal growth depends on the ability of artificial intelligence to learn and adapt, and the growth of artificial intelligence has accelerated due to the post-COVID-19 global environment emphasising digital technology. This has changed how people communicate and the cost of doing so. In the global world, it is now possible to communicate quickly and relatively cheaply with others regardless of geography. Coupled with this change in information communications, technology has been increased computing power that enables more stronger data analysis (Giuggioli and Pellegrini, 2023). Artificial intelligence allows applications and devices to act like a human as they learn new information that leads them to respond in a similar way to humans. This ability is transforming society and altering societal practices.

Artificial intelligence is a way to reduce human interaction by enabling computers to do human thinking and applies in many different industries as the idea is the same in terms of using computers for human performance (Chalmers et al., 2021). By saving time and energy on tasks that were done by humans, it can bring about productivity gains. Despite its usefulness, there are mixed feelings about artificial intelligence depending on who is speaking and their perception of the technology (Etzioni and Etzioni, 2017). It can mimic and now exceed human intelligence that is changing the way technology is used in society. By using computing analytics, it can study behaviour and foresee change. Artificial intelligence is trained on data so there are weaknesses as to its ability based on the information received (Basu et al., 2025). The usefulness of artificial intelligence is in understanding patterns and information that could be missed by humans. By decoding patterns, artificial intelligence can empower machines to make data-based decisions.

This article addresses a gap in research on artificial intelligences usefulness for society policy that emphasises the complexity between emerging technology and societal debates.

A narrative review of existing literature on artificial intelligence and social policy was conducted in order to assess the current state of the literature by critically analysing the current knowledge in an objective way. This included a non-quantitative approach in terms of searching for articles that included discussion on sociology, artificial intelligence and/or digital technology innovation. The focus in the review approach was to identify trends and broad thematic interpretations about the linkage between artificial intelligence and social policy. The review focused on journal articles, books and book chapters but relevant internet references were also consulted. By stressing how artificial intelligence enhances social policy making a purposeful research, discussion can result. This can guide social policy makers in utilising artificial intelligence for beneficial outcomes. Understanding the complexity of artificial intelligence provides social policy makers with insights that can foster better engagement. This allows them to go beyond the current mindset about artificial intelligence to project future usages. Social policy makers can draw inspiration from this approach to think in a meaningful way about artificial intelligence.

The history of artificial intelligence is complex and difficult to describe due to a number of pivotal events shaping the field. Most significantly, Alan Turing devised the Turing test to determine if computers were able to have human intelligence. Alan Turing published an article titled “computing machinery and intelligence” in 1950 that specifically linked computers to intelligence systems. This seminal thinking produced interest in future computing applications. In 1956 John McCarthy at a workshop at Dartmouth University recognised the ability of machines to potentially be intelligent and started to use the term artificial intelligence. The initial idea of computers having intelligence was that they could engage in a process of heuristic problem solving. This meant computers could solve problems based on a set of guiding principles. By doing so, computers could search potential suggestions and then be programmed to find the best results. Early artificial intelligence programs were based on games such as chess and checkers or experiences including shopping. This includes computer programs such as Shopper by Anthony Oettinger in 1951 that replicated a shopping experience.

In 1966 a computer program called ELIZA was developed by Joseph Weizenbaum. It could use natural language commands enabling it to engage with humans in discussions. By enabling human conversation, ELIZA is considered one of the first chatbots. In 1966 at the Stanford Research Initiative, a mobile robot called Shakey was developed. This technology involved usage of artificial intelligence in enabling techniques to be used in terms of object manipulation. In 1969 when humans first landed on the moon, there was an increased interest in science fiction and artificial intelligence. Organisations like NASA utilised increased computing power for space travel, and this spurred a focus on new technologies.

In the 1970s and 1980s, there was a slowness in the amount of artificial intelligence discoveries. This could be due to the introduction of the desktop computer and mobile music devices such as the Walkman becoming popular. This time period is called the artificial winter as it resulted in few radical innovations as developers were in state of slumber. In 1974 a report by James Lighthill was released that said that artificial intelligence research has under delivered on promises, which resulted in funding cuts. In the 1980s more attention was placed on the use of computers for tasks such as writing documents and doing spreadsheets, which lead to more at home usage. In 1996 the IBM developed chess program called Deep Blue that won a match against a human, which lead to increased interest in artificial intelligence. In 2000, a social robot called Kismet was developed that popularised the use of technology for human processes. In 2004 NASA sent two rovers called Spirit and Opportunity to Mars. This made people realise that artificial intelligence was useful in uninhabitable human environments. In 2011 a computer system called Watson Deep QA was developed that could respond to questions. In 2011 Apple developed a virtual assistant called Siri that was followed in 2014 by Amazon having a virtual assistant called Alexa. In 2015 Hanson Robotics debuted a human-like robot called Sophia. In 2020 Open AI developed GPT-3 which was a large language model that was trained on inputted data that could learn and evolve. In 2022 ChatGPT was released, which was a chatbot that users could ask to do a variety of tasks. In 2023 Google developed a chatbot called Bard. The rapid increase in artificial intelligence technology followed the 2020–2021 COVID-19 pandemic in which more people began using digital technology. More recently in 2022–2025, increased data centres have been built around the world in order that artificial intelligence can be used. This has created some controversy especially from sustainability experts who commented about the amount of energy required to do large-scale computing tasks. In 2024 there has been more emphasis on artificial intelligent agents who can do multi-tasks. This has led to interest on artificial intelligent–ready data that encourage better quality information to be used. To do this, there has been more investment in artificial intelligence infrastructure and regulatory enforcement. This involves labelling content generated from artificial intelligence in a clear and easy way.

The most significant change in society is artificial intelligence, and the business landscape is altering due to this ever-changing technology (Chan and Choi, 2025). The defining characteristic of artificial intelligence is novelty in terms of how it is altering business culture. The peculiarities of artificial intelligence in terms of its newness permeate business interactions. To evolve as business entities, there needs to be an integration of artificial intelligence as a strategic priority. The desire for businesses to compete drives the intention to use artificial intelligence. This can be different as the intention to increase performance may conflict with actual ways artificial intelligence can be used. A mindset open to change should be encouraged in businesses in order to foster innovation. This is valuable in terms of thinking originally and superseding competitors. Whilst artificial intelligence can lead to competitive advantages, the usage raises questions about social policy regarding ethics (Bostrom and Yudkowsky, 2018). This is due to artificial intelligence being viewed as a threat to current business models.

The adoption of any form of new technology is dependent on its usefulness, cost and ease of use. Artificial intelligence can come in a variety of forms with some less costly than others (Narang et al., 2025). The fear of change can be another reason for the slow adoption process due to legacy constraints. This means artificial intelligence usage may be concentrated in larger firms or government entities with the capacity to pay (Gupta et al., 2023). Smaller firms and individuals can and still do use artificial intelligence but in different ways. This creates inequalities regarding artificial intelligence usage that tends to be more significant in different geographic regions. Previous waves of technological innovation tended to create level playing fields when the technology was deemed as a necessity as what happened with internet usage.

The successful adoption of artificial intelligence depends on a combination of social, technological and behavioural factors (Naz and Kashif, 2025). Social factors such as usefulness and influence can be a facilitating condition. Technology reasons including performance expectations and need for interaction influence adoption rates. Behavioural factors such as personal interest reflect cognitive reasons for adoption. The journey towards artificial intelligence adoption can be complicated due to deficiencies in digital skills (Roudy, 2022). This means businesses need to acquire new skills in order to modify their business methods. This will help in navigating challenging times and respond to crises in an effective way. Due to the change in operational methods caused by artificial intelligence, many businesses remain cautious about its usage due to their scepticism and lack of knowledge about artificial intelligence capabilities (Sahut and Laroche, 2025).

Artificial intelligence is part of the next industrial revolution, which is changing the way people work and live in society (Saura and Bužinskienė, 2025). It poses new challenges that have consequences for social relationships and human well-being. As more people use artificial intelligence, it will become harder for them to acquire their own information as it is easier to rely on computer systems (Singh et al., 2026). This is worrying as some of the information on artificial intelligence may be false and/or misleading. Individuals may assume that the responses to their questions are correct rather than assuming they are only based on inputted information. As a result, people will have to navigate an increasingly digital world that uses artificial intelligence (Uriarte et al., 2026).

Intelligence is a gift and a characteristic of humans but new technological advances in the form of artificial intelligence question the opportunity of new forms of wisdom (Zeb et al., 2025). Intelligence beings, whether human or artificial, need to be responsible in how they use knowledge. Those with reasoning abilities should be responsible in its usage. When intelligence is used in the right way, it should provide goodness to society. Artificial intelligence is a human creation and can do things increasingly at faster speeds than humans (Ratten and Jones, 2023). This raises ethical questions about human safety and has prompted many to delve into the moral issues associated with artificial intelligence. Artificial intelligence is a significant technology that has ethical implications and should be directed towards social goals (Roundy and Asllani, 2024).

Intelligence is a concept that varies from person to person due to innate and environmental factors. Most artificial intelligence relies on statistical inferences based on inputted information rather than intuition (Truong et al., 2023). This is a major difference between human and artificial intelligence. Computing technology relies on patterns rather than outliers to infer meanings, which means it is good at answering questions but not so good at logical deduction. Knowledge can be acquired in different ways, and this includes through association and communication. Humans learn about things when they recognise it as useful information. Artificial intelligence involves the science of using knowledge through computing systems. It involves designing intelligence agents that can take action from data. This means it has the capacity to perform functions similar to a human. Artificial intelligence is increasingly getting more sophisticated as it induces human behaviour such as judgement and thought. By imitating human behaviour, the goal of artificial intelligence is to engage in critical appraisal in order to get the best results. This means selecting the best opinions and then contemplating how to respond. Learning can be through an experimentation process by trying different things. This means a trial-and-error method is done in order to store information about best approaches. Other forms of learning can involve categorisation in which information is stored into categories that are then used for new solutions.

Advanced technologies have paved the way for more complex social policy ideas to emerge. A natural progression of technology is artificial intelligence that builds on the foundations of previous innovations (Winkler et al., 2023). Artificial intelligence in a social policy setting refers to the usage of computational systems to analyse and act upon datasets to provide responses. It can integrate real-time information with stored data to translate into insights. The key issues social policy makers are concerned about with artificial intelligence involve unfair advantages from technology. This means data might be compromised that can lead to incorrect results. Data can vary based on how it is acquired so this can lead to potential biases. To deal with these issues, the data should be tested and added to in order to ensure its robustness. There has been much said about the potential dangers artificial intelligence poses to society. Much of this worry is due to the fear of the unknown. This is natural as with anything new, there is a sense of trepidation as to how it will be integrated into society. Artificial intelligence has limitless potential in enabling society to work better but there are questions about how to best use the technology.

Human centric artificial intelligence involves placing a personal emphasis on human–computer interaction (Upadhyay et al., 2022). It is an ethical approach to artificial intelligence as it recognises its immense potential. To ensure artificial intelligence is used to serve humanity, new policies need to be developed. This will ensure that the dignity of each person is considered in terms of using the technology. It is up to social policy advisors to create good conditions for artificial intelligence. This will enable the excitement surrounding artificial intelligence to align with societal benefits.

The main types of artificial intelligence can be classified as narrow, general and super. Narrow artificial intelligence refers to applications meant to solve specific tasks. Computers are programmed to analyse data and then formulate a response. They are not self-aware and show no intuition in doing these tasks. The next stage of artificial intelligence is general that is a bit more complicated as it generates more ideas form the inputted data (Redondo-Rodríguez et al., 2025). It is similar to humans in terms of being able to respond to information. Super forms of artificial intelligence are extending human behaviour, which means the computer system can make decisions and analyse environmental factors (Ratten, 2024a).

Artificial intelligence can be in a simple form in terms of responding to inputs based on programmed material, which means the artificial intelligence does not learn from past experiences nor evolves (Gofman and Jin, 2024). The next type of artificial intelligence is more sophisticated and can learn from data to make changes. Generative artificial intelligence has recently been adopted by many in programs such as ChatGPT as it enables more sophisticated outputs (Nenni et al., 2025). This evolves training machines from data and relationship information in order to assess situations. Self-aware artificial intelligence can decipher feelings and beliefs of others as a form of emotional intelligence. This more advanced artificial intelligence not only is able to possess a state of consciousness but also does complicated analysis (Karunakaran et al., 2025). Intelligence that is natural is deemed biological as it is based on neural sensors that provide a knowledge system. In general, knowledge provides a way of collating and then interpreting information. It involves collecting data through education and experience that can be used at future points of time. This enables information from different sources to be stored in order to help understand future events. The information known by an individual is based on facts and perceptions. This can include practical understandings about something.

The main subfields of artificial intelligence involve natural language processing, machine learning and deep learning (Hillebrand et al., 2025). Natural language processing is a way for computer systems to use human language data to make decisions. The information conveyed via voice or text is an important way for humans to decipher meaning from data. Machine learning is a way computer systems utilise data to do new tasks. By doing so they engage in a pattern identification process that enables information to be altered into knowledge that is then stored (Raisch and Fomina, 2025). This allows machines to adapt and progress thereby engaging in a learning process. Deep learning goes further than machine learning as machines can learn but also make decisions. This facilitates a multi-layer neural network approach in which data is computed into actions.

Society’s relationship with artificial intelligence is concerning as issues such as issues such as integrity need to be considered (Hossain et al., 2025). This means whilst the general purpose of artificial intelligence is technological innovation it should be managed with care. Society should use technological innovation but at the same time be cautious with how it is used.

Every historical technological innovation has been somewhat surprising and taken time to be fully implemented within society (Ramaul et al., 2026). The logic behind technological innovation is to advance science but it has embedded shortcomings that need to be overcome. This means the perception of usefulness of artificial intelligence can be based on management need (Hagtvedt et al., 2025). Digital technology can be structured in a way to persuade others in that they are often developed by for-profit businesses. This means society is exposed to marketing that might differ to reality. As humans we are interested in science, and our social bonds can be altered based on new types of technology. In this way, perceptions about artificial intelligence are spread not only by others in our close social circle but also by messages received through online and other sources. As a consequence, human relationships with truth can be altered, but when technology and human reasoning act together, it provides better results.

Artificial intelligence is a technology created by humans so it should also be managed by humans and not left to its own devices. The myth of artificial intelligence overtaking humans is worrying as it is a technology tool (Li et al., 2025). Artificial intelligence is a remarkable and notable technology but it was created by humans who have a responsibility for its moral usage. So a more societal ecosystem approach should be taken into account in terms how the technology is used so it considers simultaneously issues of science, ethics and innovation. This will ensure more equity in decisions regarding artificial intelligence including issues of risk and power (Vecchietti et al., 2025). These issues are inseparable and need to be considered tighter. This means societal discernment is required in analysing artificial intelligence. Societal advancement is needed, but it must be accompanied by human progress, with the common societal good kept at the forefront.

Social policies regarding artificial intelligence should include imposing age, time and topic restrictions. This will enable better monitoring about concerns regarding artificial intelligence such as dependency and ethics. Artificial intelligence can hook users and make them dependent on them for decisions. This is problematic when the technology might not be working properly. The appeal of artificial intelligence is that it is human like and validating but should not be a replacement for critical thinking.

Artificial intelligence and its associated technologies represent the future of sociology and social policy discussions. There is an increasing usage of artificial intelligence in policy debates that reflects its emerging technology status in society. This will impact humans and other entities in a variety of ways: some good but some bad. This means in the coming years, more attention needs to be placed on regulatory and legislative principles regarding its implementation into society. By integrating artificial intelligence elements into social policy, there will hopefully be more co-creation of value. This enhancement of the social policy experience brings opportunities and challenges. Work design is being redefined by artificial intelligence that can result in efficiency improvements. This may be a challenge to those whose jobs are replaced but can improve task performance. Repetitive tasks required by social policy workers such as monitoring data can be replaced by machine learning. This enables time to be better spent on enjoyable activities. Social policy planners not only must show a willingness to utilise artificial intelligence but also be mindful about its harmful effects. Increasing demand for personalised social policy has led to an interest in artificial intelligence. It can be utilised as a form of ambient intelligence to understand atmospheric effects on social policy. This enables more monitoring of before, during and after a social policy has been implemented. Preference regarding social policy can be analysed to create new value which can initiative corrective actions.

The continuous integration and improvement of artificial intelligence will lead to smarter social policies. Future applications will be characterised by artificial intelligence analytics and infrastructure, thereby extending the reach of people’s abilities to utilise artificial intelligence in real time. Future research needs to analyse both the bright and dark side of artificial intelligence in order to fully understand how it is affecting social policy.

Artificial intelligence promises advances in efficiency regarding the implementation of social policy; whether this is true only time will tell as there is still much uncertainty about its usage. Unfavourable outcomes may result from the misuse of artificial intelligence when in reality humans are needed. This means more research is required on human–computer interaction and how this influences decision-making (Wexler and Oberlander, 2023). Fear of artificial intelligence may impede the process as social policy planners are reluctant to adopt the technology.

Technology luddites do exist in society, and time may be needed to make people more comfortable. A reduction in human-to-human interactions from engaging in social policy decisions may affect the process. Research is required on how artificial intelligence is changing policy makers’ roles and tasks. This includes value co-creation as well as value co-destruction regarding artificial intelligence and social policy. More trust is required in the process of integrating artificial intelligence; so future research avenues need to focus on this linkage (Vicsek, 2021). Work regarding how to facilitate trust in artificial intelligence is required.

Social policy often involves sensitive and controversial topics that need to be handled in a careful manner. This means future research requires more attention placed on how to make sure information regarding social policies is kept secure. As artificial intelligence programs may require the use of large data sources, it is important that the information is not shared (Vicsek et al., 2025). This may be hard to do when the artificial intelligence technologies necessitate the collection of data in order to learn (Sanusi et al., 2026). How to do this is complex so future research work is needed on the ethics of data collection, retention and reuse. This will help alleviate concerns regarding personal data and misuse of information. The rapid pace of artificial intelligence innovations may exceed how quickly social policy planners keep up with these changes (Silva, 2026). So, more research into proactive strategies regarding artificial intelligence usage in social policy is needed.

People are getting used to artificial intelligence, so the next step is using it in a smarter way in order to yield better results (Ratten, 2024b). This means researchers need to step up to the task and design new research studies in order to obtain more insights. As recent advances in artificial intelligence are focussing on interconnectivity, research about innovative approaches is required. This could include linking in with technologists to forecast what artificial intelligence usage will look like in 10 years’ time. This will provide further information about new technology practices regarding artificial intelligence (Nazeer and Ahmad, 2025). Smart artificial intelligence emphasises better usage of technology in a holistic sense. This means the user experience is enhanced through providing user profiling. Context-related usage of artificial intelligence requires further study in order to offer customised services. This will assist social policy planners with tools to make recommendations about artificial intelligence to others. Researchers should develop a great variety of studies about artificial intelligence in order to extend existing studies on technology innovation. Research on the conceptualisation of artificial intelligence in policy discussion is needed in order to study its usefulness (Paglieri et al., 2025). This will benefit those who derive advantages from social policies as well as the social policy makers.

This article has several theoretical implications for the academic and policy fields of artificial intelligence (Liu, 2021). It shows how new theory is required to deal with the quickness of new technological innovations such as artificial intelligence. Technological advancement is part of life and needed for entrepreneurial gain. Artificial intelligence at the moment is still not the same as human intelligence as it is based on individual capability (Kumar and Suthar, 2026). This means human intelligence differs between people and is evident in different ways. A machine can never fully replicate this intelligence as much of it is unknown. Whilst artificial intelligence enables tasks to be performed, it cannot think like a human. Whether thinking can be done by a computer is subject to debate and is a philosophical idea. Thinking is based on many factors including tangible and intangible knowledge. The human mind evaluates social interactions and context as well as linking them to experience. This means a range of emotional responses including sensory input are utilised in the thinking process. There should be a cautionary tone placed on unsupervised usage of artificial intelligence (Larasati et al., 2023). This means respecting human dignity in the way it is integrated into society.

Artificial intelligence–driven technologies are enabling the creation of social policy that bypasses human input (Kumar and Sahu, 2026). This means it can be easier to write social policy without paying labour rates but it can lead to lower quality social policies. Artificial intelligent–made social policy can bias views whilst not taking into account other perspectives (Garcia and Kwok, 2025). The data inputted into machines used by artificial intelligence can be based on stereotypes that marginalise other cultural views. This can reinforce views whilst not taking into account diversity.

Whilst there are many existing theories about technological adoption such as social cognitive theory and the theory of reasoned action, new theories are required to keep up to date with new developments. This will ensure artificial intelligence is integrated within existing theories but also helps to produce new theories (Feher and Veres, 2023). The question is whether existing theories can include artificial intelligence or if it requires new theories. In relation to social policy, artificial intelligence has significant implications, so it is worthy of new theory development.

The practical implications of integrating artificial intelligence into social policy discourse are immense due to its usefulness. Artificial intelligence is a life-altering technology that is bringing about discussions about the link between technology and ethics (Gama and Magistretti, 2025). At risk are human connections and relationships that are important parts of society. This means humans need to be thoughtful about how the technology is shaping human lives and whether different choices regarding its usages can be implemented in society (Boyd and Holton, 2018). Digital ethicists are suggesting ways technology can include more human thought in order to shape a better society. The constant usages of artificial intelligence have led to people thinking about it as a digital person in their lives but in reality, it is a machine (Abbas, 2025). It is nonsense to equate artificial intelligence with human empathy but it is a possibility. True empathy is part of the relational sphere in requiring the lived experiences of others. This skill can be replicated in some ways but does not recognise human qualities such as uniqueness. Other feelings such as love are hard to replicate because it is based on experiences. Digital philosophers can debate whether technology can experience feelings of empathy and love as a machine is made by humans and not born like humans. This means there needs to be a distinction between artificial intelligence and humans. This can be difficult when there are more people who believe they can express similar feelings.

Of primary importance is using artificial intelligence in innovative ways to better understand societal trends (Al Yakin et al., 2026). This will help predict patterns that social policy advisors can monitor. By doing so, there needs to be more collaboration between technologists and social policy workers. This will ensure that the basic requirements of social policy are met but future considerations are also considered by social policy makers. The promotion of good social policy is needed to ensure it meets different contextual requirements from artificial intelligence usage. This means public institutions should consider the implications of artificial intelligence on how people interact in society. Academics can then apply these findings in future studies regarding how artificial intelligence has changed policy in terms of improvements. This will help in identifying best practice and decisions regarding future research avenues regarding artificial intelligence, sociology and social policy.

Artificial intelligence is exposing the inability to make progressive social policy. The realities of the digital economy have led to increased usages of technology to write, create and analyse social policy. This has accelerated the need to be tech-savvy by social policy makers. These changes are not occurring in a vacuum but are rather coinciding with bigger workplace changes. Current social policy recognises the need to advance policy discussions as part of overall development. Yet the usage of artificial intelligence has not been implemented quickly in a social policy setting. This makes social policy outdated and unable to adapt to further technological change. The disruptions caused by artificial intelligence are requiring more creative usages of social policy; so this article is amongst the first to explicitly link the areas together. In summary, this article seeks to expand our understanding of artificial intelligence and social policy by taking a narrative review approach to identify key themes of analysis. This enables a contribution to the broader realm of how artificial intelligence can serve as a strategy for promoting social policy initiatives.