Human Checklist

Key features and attributes to make sure your AI agent serves humans effectively, making interactions both efficient and safe.

Does your agent...
use each human's preferred language?
offer the user's preferred modality? E.g. text, speech, video...
offer the user's preferred channel? E.g. voice, SMS, Messenger, email...
provide for users with disabilities?
personalize interactions based on history, context and knowledge of the user?
adapt to different user levels of technical or domain expertise?
encourage and enhance collaboration between humans?
act as a collaborative equal rather than a subordinate tool
avoid unnecessary complexity and jargon?
get confirmation when making critical decisions?
detect misunderstandings, then paraphrase or ask questions to rectify them?
provide easy access to human support?
explain its non-obvious decisions and actions?
filter content that may cause harm or reputational damage?
log interactions in aggregate and check itself for bias?
adapt to human emotional and social cues?
reinforce user confidence and autonomy?
detect signs of signs of harm or crisis and escalate where appropriate?

Satisfaction Risk: 100

Humans are expecting more and more from the AI agents they interact with. They won't be satisfied or productive unless your agent implements most of the features on this checklist. Check the boxes to answer the above questions and get an assessment of your agent's satisfaction risk.

AI Agent capabilities are evolving quickly - particularly their modes of communication and their emotional intelligence. Be sure to check this list regularly for new knowledge.