Understanding “Data” in AI, Algorithmic and Autonomous Systems – A Concept Framework
INTRODUCTION
AAA systems are socio-technical
systems. Humans are embedded in
the AAA System from the beginning
of the process all the way to the
end. Whether through the use of
Personal Data or outcomes, AAA
Systems run the risk of negative
impacts to humans and/or our
humanity. Through differentiation
and precision, ForHumanity aims to
integrate the law (EU Artificial
Intelligence Act/GDPR) and industry
practice with Independent Audit of AI
Systems to ensure Governance,
Oversight and Accountability of every
moment and discrete occurrence of
“data”.
Data transits across the algorithmic
lifecycle and is ever present along the
way - existing both before the AAA
System is designed and after the AAA
system is decommissioned. It exists
in many forms, frequently
concurrently and thus the conundrum
of taxonomy. The concurrence and
multidimensional nature of data
demands clarity and specificity.
Precise understanding of Data is
necessary because when misused, the
potential for harms or negative
outcomes for natural persons
exists. When well defined and well
understood, we can assess, analyze,
evaluate and maximize risk
mitigations associated with data
across the entire lifecycle of AAA
Systems.
