Challenges to Grassroots Organization Engagement with AI Policy

Collection of recent arXiv papers (June 19, 2026) examining governance, risks, and societal impacts of advanced AI. Key themes: challenges for grassroots and marginalized communities to engage in AI policymaking; the need for proportional evaluation frameworks for open-weight models (OWMs) to assess misuse and robustness; methods and interpretive cautions for evaluating biological capabilities of agentic AI (biosecurity implications); evidence of LLM-driven cyber-espionage and unequal access to frontier models (geopolitical exclusion, especially Africa); environmental costs of image-generation