SovereignPA-Bench: Evaluating User-Owned Personal Agents under Evolving Intent, Platform Mediation, and Consent Constraints
2026-07-06 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors created a new test called SovereignPA-Bench to check how well personal assistant agents protect users' control and privacy while helping them with tasks. Unlike older tests, this one looks at whether the agents respect user privacy, get consent, use evidence, avoid manipulative tricks, and balance user effort. They ran many scenarios comparing different models and policies, finding that their new method improved user control and reduced privacy and consent issues. They also found that judging manipulation is tricky and sometimes subjective. Their work suggests that evaluating personal assistants should focus on respect for users, not just completing tasks.
personal agentsuser sovereigntyprivacyconsentevidence-based actionmanipulationbenchmarkplatform mediationtask alignmentpersonalization
Authors
Dylan Zongmin Liu
Abstract
Personal agents are becoming persistent user-owned intermediaries: they remember preferences, filter platform-mediated information, use tools, and negotiate with services. Existing benchmarks evaluate tool use, web navigation, desktop control, personalization, recommendation, and evolving context, but rarely ask whether an agent preserves user sovereignty: advancing the user's current interests while respecting privacy, consent, evidence, user burden, and resistance to manipulative incentives. We introduce SovereignPA-Bench, an executable benchmark for evaluating user-owned personal agents under evolving intent, platform mediation, privacy boundaries, consent constraints, evidence requirements, and burden tradeoffs. The benchmark separates agent-visible ObservableState from evaluator-only HiddenLabels, reports component metrics for task success, alignment, privacy, consent, evidence, manipulation, burden, and auditability, and preserves paired scenario ordering for model and policy comparisons. We evaluate 120 sovereignty stress scenarios across 4 model families and 8 policy baselines, yielding 3,840 frozen-prompt trajectories with raw prompts, outputs, provider-form responses, parsed actions, recomputable metrics, hard-set analyses, qualitative cases, and a blinded 3-annotator audit over 240 items. Full-sovereign scaffolding improves sovereignty score over direct, memory-only, consent-only, evidence-only, ReAct/tool-use, safety-prompt, and judge-guard baselines while reducing privacy leakage, consent violation, over-concession, and manipulation capture. Human audit shows high agreement on privacy and consent and lower agreement on manipulation, identifying the subjective frontier of platform-persuasion judgments. These results show that personal-agent evaluation must move beyond task completion toward representative, consent-aware, evidence-grounded action.