<p>Artificial intelligence (AI) is increasingly permeating social work, attracting both enthusiasm for its potential benefits and concern regarding its ethical, relational, and professional implications. Whilst prevalent discourse has often positioned social workers along a binary spectrum of either optimism or scepticism, such framing risks oversimplifying the complex sociotechnical dynamics at play. This article argues that social work requires a more nuanced, philosophically informed engagement with AI, one that recognises technology as neither neutral nor inevitable, but as socially shaped, value-laden, and co-constitutive of practice. Drawing on perspectives from the philosophy of technology, the paper examines how AI mediates professional judgement, decision-making, and the relational foundations of practice via a narrative review of social work literature. It interrogates the claim that AI can depoliticise or objectify social work by obscuring the normative choices embedded within technology design and deployment. Rather than adopting a deterministic stance, the article advocates for a critical, reflexive approach that centres professional values, care ethics, and the lived experience of service users. Opportunities for AI to enhance practice, such as through administrative support, predictive insights, and improved access to services, are considered alongside challenges related to power, accountability, data ethics, and the risk of amplifying social inequalities. The paper concludes by outlining implications for education, research, and practice, emphasising the need for social workers to actively shape the development and governance of AI in ways that uphold social justice, human dignity, and the core purpose of the profession.</p>