The intensified digital divide: Comprehending GenAI

In the swiftly evolving digital landscape, the advent of generative artificial intelligence (GenAI) is heralding unprecedented changes in how we interact, work, and innovate. However, this technological renaissance brings to the fore a critical yet often overlooked aspect: widening the existing digital divide. As we stand on the brink of this new era, it's imperative to explore how GenAI literacy and subsequently associated literacies — such as prompt literacy (Maloy & Gattupalli, 2024) and AI literacy (Long & Magerko, 2020) — are becoming pivotal elements of the digital divide, deepening the chasm between the digitally privileged, i.e. those who can comprehend their logic, and those on the periphery.

Beyond access to understanding

Since the late 1990s, the concept of the digital divide — referring to the disparity between those with and without access to computers and the internet — emerged as a pivotal concern in research and politics in relation to the advancement of new media (Van Dijk, 2006; NTIA & U.S. Department Of Commerce, 1999). Over the years, a broader understanding of the digital divide has evolved, recognising that it extends beyond physical access to technology to include the effective use of these technologies (Gunkel, 2003). In this context, many scholars have envisioned digital literacy as an aspect of the digital divide (see, for example, Chetty et al., 2018). Moreover, studies suggest that digital disparities significantly reflect broader socioeconomic inequalities, encompassing gaps in access, engagement, and effective utilization of digital technologies (Vassilakopoulou & Hustad, 2021). Thus, we can notice a transition from focusing on physical access to an emphasis on digital skills and their application (Van Dijk, 2006). With the recent widespread use of AI and its language models, we are now witnessing the intensification and broadening of digital literacy issues (Kindarji & Wong, 2023; Wang et al., 2023).

While previous studies acknowledge the expanding digital divide and the potential of AI, and accordingly, GenAI, to exacerbate existing inequalities, they do not delve deeply into the specific challenges posed by the need to understand, create with, and control GenAI technologies. In this article, I argue that GenAI is forging a new digital divide concerning not only who can log on or possesses technical digital skills but also who can comprehend, create, or control AI technologies. In this regard, it is inequalities in knowledge (comprehension of algorithms and skills to use them) which are one of the rising digital inequalities, alongside databases and treatment inequalities (Ragnedda, 2020). My argument builds on previous work on "AI literacy," understood as the essential competencies for critically assessing, effectively interacting with and utilizing AI in various settings, covering understanding AI systems, critically evaluating AI uses, and weighing AI's risks and opportunities (Long & Magerko, 2020).

Hence, GenAI literacy involves understanding how generative models are created and trained, how they generate new content, the limitations and potentials of this technology, and the unique ethical considerations they entail, such as authorship, authenticity, and misinformation. GenAI literacy encompasses a range of skills, from basic awareness of AI principles to advanced capabilities in comprehending, developing, and managing AI systems. This literacy spectrum not only demands an understanding of how AI operates but also requires the comfort, (see Li et al., 2024), ​​ability, and perhaps trust, to interact with and guide AI systems effectively. The deepening of such literacy and the attainment of its related skills mirrors and exacerbates the current digital literacy divides, primarily due to unequal access to necessary technological resources and educational opportunities. Individuals from lower socio-economic backgrounds may find it challenging to access cutting-edge AI technologies or acquire the knowledge to utilise them effectively. Moreover, the complexity of these technologies and their multifaceted ethical implications introduce additional layers of disparity. Those without the means to engage in higher education or specialised training might struggle to keep pace with AI advancements, widening the gap between AI experts and the general populace. Consequently, the expansion of GenAI literacy underlines the need for inclusive educational initiatives, capacity building for critical thinking, and equitable access to AI resources to mitigate these divides and ensure that the benefits of AI advancements are accessible to all.

The compounded impact of GenAI on society

The implications of the broadening digital divide due to GenAI are profound and far-reaching across different sectors (Capraro et al., 2023). For example, in education, the gap widens not merely due to the absence of AI-enhanced tools but also because of the lack of structured programs for AI literacy (see Duran & Ermiş, 2024; McDonald et al., 2024). In the job market, the absence of GenAI skills translates to more than just a disadvantage; it actively excludes individuals from the hiring process. Many companies now use AI for resume screening, prioritising candidates who demonstrate fluency in AI-related competencies (ResumeMent, 2023; Yu, 2023). For businesses and entrepreneurs, particularly in economically marginalised communities, the struggle to integrate AI technologies stems from a complex interplay of factors including limited financial resources, a dearth of technical expertise, and inadequate access to reliable internet services. This lack of AI adoption not only stifles their operational efficiency and market competitiveness but also limits their ability to engage in data-driven decision-making, a critical component of contemporary business strategies (see Adams, 2023).

These manifestations of the GenAI divide highlight a critical reality: the barriers to accessing and leveraging AI technologies are deeply intertwined with existing socioeconomic structures. Without targeted interventions to enhance AI literacy, democratise access to AI tools, and foster an inclusive digital ecosystem, the potential of GenAI to exacerbate social and economic inequalities remains a pressing concern (Chaudhuri, 2023). The essence of the issue lies in the compound intersection of economic, educational, and infrastructural inequities with the rapid advancement and integration of GenAI technologies across sectors. As these technologies become entwined with everyday functions and professional requirements, the divide deepens, propagating existing socio-economic disparities and gives rise to new forms of inequality (Mandapuram et al., 2018).

Moving forward: Bridging the new digital divide in the age of Generative AI

As we navigate the complexities of this digital transformation, it's crucial not just to celebrate advancements but also to critically examine who might be left behind. By fostering a collective dialogue on GenAI literacy and the new digital divide, we can explore inclusive strategies that ensure everyone has a seat at the table. This is not just about equitable access to technology but about empowering all individuals with the knowledge and skills to thrive in an AI-driven world (GGI Insights, 2024).

The journey towards bridging the new digital divide is complex and fraught with challenges. Yet it presents an opportunity to redefine digital inclusion and ensure that the future of AI benefits humanity in its entirety, without exception. Accordingly, it is imperative to prepare individuals for a future where AI is omnipresent. Making GenAI tools and resources accessible and affordable is vital, as is fostering collaborations between public and private sectors to provide the necessary resources and infrastructure for widespread GenAI education, comprehension, and adoption (Veitch, 2023). Addressing the ethical dimension of GenAI development is equally critical. The inherent biases in AI, issues of academic integrity, privacy, data security, accessibility, and equity are ethical considerations that society and its institutions must navigate as they integrate AI technologies. These ethical challenges require collaborative approaches among diverse stakeholders to ensure equity, transparency, and responsible use.

Moreover, to tackle the AI digital divide, a broad approach involving economic redistribution, universal basic income, extensive training programs, state-supported AI research, and enhancement of digital infrastructure and skills is crucial, demanding collaborative efforts from governments, businesses, and civil societies to secure a fair and inclusive future driven by AI (Rana & Verhoeven, 2023). There need to be regulations that properly and directly address AI digital literacy issues and the digital divide. Thus far, policy has failed to keep pace with these societal challenges. For example, the first legal framework on AI, the EU AI Act, does not explicitly address digital literacy and the digital divide. The Act primarily focuses on the regulation of AI technology rather than its societal effects, addressing literacy only implicitly, through various provisions and initiatives designed to enhance transparency, accountability, and public engagement with AI technologies (see Future of Life Institute, n.d.). The focus on technology regulation without engaging in the societal impact of said technology is a weakness, one which future AI policy must be wary of unthinkingly replicating.We must approach the new digital divide with a nuanced understanding that accessibility does not solely equate to the provision, or even regulation of technology itself. Rather, it involves crafting policies that ensure not only equitable access but also equitable comprehension of the full spectrum of digital resources, fostering an environment where diverse groups can build digital literacy and proficiency. It is not enough to close the connectivity gap; we must also close the comprehension, literacy, and proficiency gaps to achieve true digital equity.