The Fractured Scale and Unreadable Data: Exceptionalism in Sexual Crime Trials vs. the Epistemic Exclusion of Chronic Violence

Modern criminal jurisprudence faces a fundamental structural paradox. On one hand, legal systems dealing withsexual offenses have increasingly embraced "narrative-based exceptionalism"—relaxing strict physical evidencerequirements in favor of victim testimony, contextual credibility, and gender sensitivity doctrines to capture offensesoccurring in secrecy. On the other hand, the rapid integration of Artificial Intelligence (AI) and predictive analyticsinto judicial decision-making enforces a stringent paradigm of data formalization, relying exclusively on quantifiable,structured digital footprints. This paper investigates the critical conflict between these two opposing trajectories. Wedemonstrate that while high-profile or politically visible sexual crime narratives manage to breach traditionalevidentiary thresholds, victimizations rooted in long-term, systemic abuse—termed "Chronic Violence"—are epistemically erased. Due to psychological mechanisms such as learned helplessness, coercive control, and trauma-induced normalization, victims of chronic violence often fail to recognize their own victimization as a discrete legal entity, thereby generating zero digital or physical data ("Unreadable Data"). As judicial AI systems automate caseassessment, risk evaluation, and evidentiary screening, they treat the absence of structured data not as a failure ofsocial capture, but as an absolute lack of legal merit. Consequently, judicial AI institutionalizes an unprecedented formof "Algorithmic Exclusion," permanently expelling non-datafied, silent victims from the domain of legal protectionwithin milliseconds. We conclude by arguing that the modernization of justice through automated algorithms risksreifying cold evidentiary formalism, making the preservation of judicial discretion, philosophical interpretation, and"algorithmic humility" imperative for preventing automated institutional violence.