Source note1 min read
Designing Loyalty: AI Agents and Conflicts of Interest
- Publication date: 2026-08-25
- Collection date: 2026-08-26
- Authors/organization: Ella Genasci Smith, Victor Y. Wu, Jennifer King / Stanford HAI
- Evidence type: university policy issue brief
- Language: en
- Source role: university/lab research
- Panel: stable
- Exact URL: https://hai.stanford.edu/policy/designing-loyalty-ai-agents-and-conflicts-of-interest
Source facts
Stanford HAI's brief examines conflicts of interest in AI agents and argues that developers and deployers should owe users a duty of loyalty, especially in higher-stakes or regulated contexts. It describes agents as systems that can perform multistep tasks, call external tools, retain information, query databases, and coordinate with other agents. The brief identifies disclosure gaps and recommends coordinated action by standards bodies, regulators, and Congress.
Analysis
This is a policy response to agentic systems gaining broader access to personal data, external tools, and delegated decision-making. It is useful as governance evidence, not as proof that a particular agent causes harm.
Uncertainty/Doubt
The recommendations are the authors' policy position; their legal and economic effects remain to be tested.
