What does it mean to use Emotion AI responsibly?
The human brain is the most complex structure in the universe. Studying emotion is not an exact science, but a blending of science and art. That’s why our emotion artificial intelligence (Emotion AI) is guided by best practices in evidence-based psychology research and the latest advances in deep machine learning.
We appreciate the sensitive nature of Emotion AI and are committed to implementing safe, equitable and transparent solutions to minimize potential harm. Our technology is therefore rooted in a person-centered approach that honors core principles.
SAFE
Ensuring that all personal data stays private and secure.
- We strictly adhere to all privacy and security standards, including the General Data Protection Regulation (GDPR).
- Our technology does not retain any Personally Identifying Information (PII) and cannot be used for micro-targeting individuals.
- Human engineers are continuously in the loop to identify potential gaps, limits or shortfalls of the technology.
EQUITABLE
Protecting individuals and communities from algorithmic discrimination.
- Our emotion algorithms have been trained on tens of millions of annotated examples collected across diverse and random backgrounds.
- We mitigate bias by accounting for linguistic variations and cultural differences by training our Emotion AI natively in the original languages without the shortcomings of translation.
- The output of our technology is data that is relative to a population over time and across topics, never in absolute terms.
TRANSPARENT
Promoting awareness using language everyone can understand.
- When it comes to Emotion AI, we believe in accountability through accessibility, working to bring our tools to people who need them most.
- We are committed to advancing the public’s understanding of Emotion AI, and identifying ways these contribute to positive outcomes.
- Our team is dedicated to demonstrating the purpose, use and value of our AI system prior to its application.