A big-tent or small-tent AI safety movement?

Newsletter · Watch/Read · (gathered 2026-10-05)

• The authors advocate for a “big-tent” AI safety movement that focuses on specific, AI-amplified systemic risks (like cyber-risk, bio-risk, and institutional decay) through diverse, practical interventions, rather than a “small-tent” approach narrowly focused on speculative AI existential risk (x-risk).
• They highlight society’s systemic “underinvestment in catastrophic risks” (e.g., pandemics, cybersecurity) and introduce “accumulative risk” to describe the gradual erosion of societal resilience. They note AI’s high public “concern” but low “salience” as a top issue, explaining why x-risk narratives gain disproportionate attention.
• The x-risk framing is deemed “actively counterproductive” because it leads to polarization, misdirects resources towards ambiguous bans, and promotes “policy-by-panic” instead of building societal resilience through common-ground policies like transparency, liability, and embedded evaluation for existing AI systems.