New Canadians participate in elections at substantially lower rates than other citizens (Stephenson, Harell, Rubenson and Loewen, 2020). Research points to several barriers: unfamiliarity with electoral procedures, practical information costs (where to vote, how to register, what identification to bring), and lack of trust in institutions newcomers have had little prior contact with. At the same time, advanced digital technologies (ADTs) are creating new opportunities to reduce these barriers by delivering accurate, accessible civic information at scale (Williamson, Hopkins, and Rubenson, 2026). This project will conduct two large-scale, pre-registered field experiments in collaboration with Elections BC and settlement-sector organizations in advance of the next British Columbia provincial election (on or before October 21, 2028). Both experiments will randomize how nonpartisan election information reaches newly naturalized citizens and measure voter turnout via linkage to the provincial voter file. The first tests whether providing practical, multilingual information about voting procedures—polling locations, mail-in ballot instructions, identification requirements— increases turnout among newcomers, and whether an interactive AI tool outperforms conventional formats. The second targets newcomers who voted for the first time in the 2024 provincial election, testing whether habit-formation outreach sustains their participation and which delivery channel—settlement organisation, peer mentor, AI chatbot, or official communication—is most effective (Gerber, Green, and Shachar 2003). Together, the experiments ask: can digital technologies reduce informational barriers to democratic participation, or does human connection remain essential? Several features of these studies set them apart. Treatment is randomly assigned, the gold standard for causal inference since in expectation the only thing differentiating treatment and control groups is receipt of treatment; the studies are intervening in real politics in collaboration with an election management body—as opposed to in surveys; and outcomes will be measured using administrative voter turnout records rather than self-reports. In the United States, voter file linkage is standard in field-experimental research on mobilization (Gerber and Green 2000; Gerber, Green, and Larimer 2008), and the ability to observe actual turnout—rather than relying on the inflated self-reports that plague survey-based research—eliminates the measurement bias that undermines even well-designed survey experiments. In Canada, however, voter files are tightly restricted and access for research purposes is extremely rare. No Canadian field experiment has combined randomized interventions targeting newcomers with administrative turnout measurement. Rubenson has maintained a continuous collaboration with Elections BC since 2011, including population-level field experiments with voter file linkage (Loewen, Rubenson, and Martin 2012). This relationship makes it possible to conduct the kind of study that is, to our knowledge, unprecedented in the Canadian context—and to do so with strong protections for participants, including formal data use agreements, research ethics board oversight, informed consent for record linkage, and de-identified outcome data.