Mobile AI Assistants and the Erosion of Human Decision-Making
Abdur Rehman · International Journal of Mobile Applications and Technologies · 2026
The rapid proliferation of mobile AI assistants — including voice-activated agents, recommendation engines, and automated decisional nudges — has fundamentally altered how individuals form preferences, evaluate choices, and reach decisions in everyday life. This paper investigates the cognitive, behavioral, and societal implications of delegating decision-making to mobile AI systems through a mixed-methods study combining systematic literature review with empirical survey analysis across 1,240 participants in six countries. Findings reveal that sustained reliance on AI-mediated recommendations is associated with a 34% reduction in decision uncertainty tolerance (willingness to defer a choice pending additional information), measurable atrophy in metacognitive self-assessment accuracy, and significant shifts in perceived personal agency. Behavioral analytics indicate that users who rely heavily on AI recommendations exhibit narrowed consideration sets, averaging 2.1 options versus 5.8 for low-reliance users, and reduced tolerance for decision uncertainty. A dual-process theoretical framework is proposed that distinguishes between efficiency-enhancing automation and autonomy-eroding automation, enabling more nuanced policy and design interventions. Results carry significant implications for AI system design, digital literacy policy, and regulatory frameworks governing AI decision-support tools in consumer contexts.