What are the tradeoffs between openness/transparency and manipulation resistance?

Proposes the “Goldilocks” equality objective for sortition algorithms that balance fairness, manipulation-robustness, and transparency when selecting representative panels for Citizens’ Assemblies. The algorithm ensures no volunteer receives too little or too much chance of selection, achieving n...

Miller, Milli, and Hardt (2020) reveal a fundamental connection between strategic classification and causal inference, distinguishing between gaming (circumventing the system) and genuine improvement. Their central argument is that designing classifiers that incentivize improvement must inevitabl...

Google’s guide defining adversarial testing as systematically evaluating ML models against malicious or inadvertently harmful input, covering explicit queries (containing policy-violating language) and implicit queries (seeming harmless but involving sensitive topics). The four-stage workflow inv...