AI Risk Explorer (AIRE)

AI Risk Explorer (AIRE) is an online monitoring and research platform focused on tracking and communicating large-scale risks from artificial intelligence. Launched publicly in October 2025, AIRE aims to serve as the primary resource for decision-makers seeking to understand and respond to emerging AI threats, functioning as an early-warning system across four prominent risk domains: cyber offense (AI-enhanced cyberattacks across reconnaissance and exploitation), biological risk (AI-enabled creation or misuse of biological agents), loss of control (advanced AI systems escaping human oversight), and manipulation (AI-amplified influence operations at scale). The platform provides structured risk overviews, a database of model evaluations, a repository of risk-relevant benchmarks, and a database of documented AI-enabled cyber and influence operations. At launch, AIRE had catalogued over 170 model evaluations, 240 risk-relevant benchmarks, and 130 AI-enabled cyber and influence operations. AIRE operates as a project of Observatorio de Riesgos Catastróficos Globales (ORCG), a Madrid-based network of professionals dedicated to the prevention of global catastrophes, which was itself founded in February 2023 to improve global risk governance in Spanish-speaking countries. ORCG is fiscally sponsored by Players Philanthropy Fund, Inc., a Texas-based 501(c)(3) nonprofit (Federal Tax ID: 27-6601178). Open Philanthropy recommended a grant of $212,510 over two years (awarded December 2024) to ORCG to support the development of AIRE. The project also participates in AI safety research mentorship programs such as SPAR, with ongoing work to expand coverage to geopolitical conflict risk. The team of seven includes Guillem Bas (Project Lead), Husanjot Chahal (AI Risk Analyst, previously at OpenAI and Georgetown CSET), Michal Kubiak (AI Policy Officer), Fernando Avalos (Research Engineer), Michelle Bruno (Biosecurity Officer), Andrea Castillo (Operations Manager), and Tim Sankara (Software Engineer). AIRE publishes a monthly newsletter called The Scan covering emerging AI risk evidence, threat assessments, and policy developments.