Artificial intelligence has the potential to promote social justice by expanding access to education, healthcare, and social services. Nonetheless, in both the short and medium-term this technology could dramatically exacerbate social imbalances — if not create them entirely — by mechanizing away jobs that previously called for low or middle levels of skill, driving people out of work. Without such initiatives as reskilling courses or fair access to AI, AI may have the effect of widening disparities between different groups rather than closing them; responsible implementation is thus imperative.

The role of artificial intelligence (AI) in the social dimension of sustainable development is indeed paradoxical, presenting both significant opportunities and serious challenges.

On one hand, AI holds immense potential to **enhance social justice and inclusion**. By improving the efficiency, accessibility, and quality of essential social services like education, healthcare, and social protection, AI can help bridge gaps that traditionally marginalized populations face. For example, AI-driven personalized learning platforms can support underserved students, telemedicine powered by AI can extend healthcare access to remote areas, and predictive analytics can help target social assistance programs more effectively. These advances contribute to reducing inequalities and promoting more equitable societies.

On the other hand, the **rapid adoption of AI and automation poses risks that may exacerbate social inequalities**, particularly through job displacement in low-skilled sectors. Many manual or routine jobs are at risk of becoming obsolete due to automation, which will disproportionately affect vulnerable workers who may lack the necessary skills or resources to adapt to new roles. Without comprehensive, inclusive public policies that support reskilling, social safety nets, and fair labor market transitions, this could deepen unemployment and income inequality, undermining social cohesion and long-term sustainable development.

Ultimately, whether AI acts as a catalyst for social justice or a driver of social disparity depends largely on **how societies manage its integration**. Proactive policymaking, inclusive education and training programs, and ethical AI governance are critical to ensure AI’s benefits are broadly shared and its risks mitigated. Fairness and transparency in AI system design, public-private partnerships for workforce development, and human-centered approaches are all examples of this. In conclusion, AI is a potent tool whose social impact depends on its responsible deployment and commitment to inclusive, sustainable development goals. It is neither inherently a problem nor a solution for social justice. When aligned with sound policies and ethical frameworks, AI can be a transformative force for social equity and cohesion. Without them, it risks reinforcing existing divides and destabilizing societies.

I agree that artificial intelligence has great potential to promote social justice especially by making information normally held behind paywalls more easily accessible. However, many problems with AI persist due to training and historical disparities perpetuated due to the training data. We are training our AI models with information that has largely been classist, racist, sexist, and, for all intents and purposes, bigoted. This bias continues this disparity even before discussing the implications of automation and job transformations due to AI.

In the grand scheme of things, AI is fairly new - but due to the historic disparities, I'm not confident that AI will help lessen social inequity UNLESS we see a major shift to include a more diverse group of programmers join the field to ensure the issues with training data biases are managed appropriately. I think AI has incredible potential to make the world better, more equitable, and more productive - but, unless we properly address the historic inequality we are training these models on, AI cannot do better for our society.

*Disclaimer: Just my two cents, obviously*

Dear Zouyene Sadek , you may find some interesting answers in this discussion room, it is related to your research question.

https://www.researchgate.net/post/Navigating_the_Impact_of_AI_on_Employment