INLAY: Preemptive, In-Context Intelligence for Casual Web Browsing

AI assistance in browsers often depends on explicit invocation, which is effective for well-defined tasks but imposes significant overhead during exploratory 'casual' browsing where intent is implicit or latent. We introduce Inlay, a preemptive INtelligence LAYer that bridges the intent-action gap herein by proactively embedding AI-driven insights and cues directly within webpages. Grounded in Information Foraging Theory, Inlay enhances 'information scents' by shifting users from high-effort query formulation to low-effort intent confirmation while preserving user agency. Our preliminary user study (N=12) provides early signals that Inlay's context-aware interventions may anticipate emergent needs and encourage deeper exploration while sustaining foraging flow. Our findings suggest that proactive augmentation can transform browsing from command-driven interaction to collaborative AI-assisted exploration.