Too Focused on Accuracy to Notice the Fallout: Towards Socially Responsible Fake News Detection

The rise of fake news is one of the most pressing threats to the digital public sphere. Artificial intelligence (AI) systems promise to fight it — but at what cost? Unlike other machine learning applications designed to optimize efficiency in low-stake domains, fake news detection operates at the core of democratic discourse, public trust and epistemic integrity. This paper begins by unpacking the core challenges that make fake news detection uniquely demanding. In response, we argue for a shift toward Socially Responsible AI (SRAI) as a more appropriate framework for addressing these complexities. We map the identified challenges onto the SRAI pyramid—functional, legal, ethical and philanthropic. Finally, we review emerging initiatives, highlight current limitations and propose future directions for developing fake news detection systems that are not only accurate but also socially accountable and publicly trustworthy.