Back to the MIT repository
3. Misinformation2 - Post-deployment

Erosion of Society

With online news feeds, both on websites and social media platforms, the news is now highly personalized for us. We risk losing a shared sense of reality, a basic solidarity.

Source: MIT AI Risk Repositorymit89

ENTITY

2 - AI

INTENT

2 - Unintentional

TIMING

2 - Post-deployment

Risk ID

mit89

Domain lineage

3. Misinformation

74 mapped risks

3.2 > Pollution of information ecosystem and loss of consensus reality

Mitigation strategy

1. Algorithmic Recalibration for Epistemic Diversity Implement systematic algorithmic interventions, such as 'serendipity features' or 'editor-curated diversity cues,' to deliberately introduce politically and ideologically heterogeneous content. This strategy aims to mitigate the reinforcement of homogenous beliefs by ensuring that recommendation systems actively prioritize exposure to a wider epistemic landscape, thereby counteracting the systemic fragmentation of consensus reality. 2. Mandated Algorithmic Transparency and Granular User Control Require platforms to provide clear, audited explanations of how content personalization algorithms function (transparency). This must be coupled with the implementation of granular user controls that allow individuals to actively adjust, or fully override, the degree of algorithmic filtering applied to their content feeds, fostering user agency and enabling self-correction against echo chambers. 3. Widespread Algorithmic and Media Literacy Education Launch comprehensive public education programs focused on building 'algorithmic literacy'—the cognitive understanding of how personalization systems and filter bubbles operate—alongside traditional media literacy. This equips the public with the critical thinking skills necessary to recognize and compensate for algorithmic bias, critically evaluate sources, and resist the emotional manipulation tactics prevalent in sensationalized, tailored content.