HOUSTON - A flurry of use cases, possibilities, and ethical concerns have arisen from the proliferation of artificial intelligence (AI). Organizations are struggling with how to use it to gain a strategic advantage, while simultaneously balancing ethical concerns that may arise. When one speaks of ethics, it is not just about the larger societal concerns - such as displacing workers or debasing human activity - but also about how organizations ensure employees do not engage in unethical behavior with AI while performing their work. When it comes to ethics for AI, organizations must empower their employees and normalize its use.

Many ethical challenges organizations will face are as yet unknown. One can speculate about how the technology might develop and the moral dilemmas that could evolve from it, but prematurely writing policies for its use will either address concerns that are misplaced or overlook problems that will seem obvious in a few years. Social media provides a valuable lesson in this regard: In its earliest days, no one could have predicted that it would lead to increased political polarization or depression.

An ethics for AI must be derived from bottom-up experience. Until companies develop a better understanding of how employees will use the technology, any attempt by upper management to create guidelines for proper use is a bit like asking an executive who has never played hockey to write rules for player safety.

What follows is a practitioner-centric approach. Practitioners are best situated to understand challenges, threats, and opportunities, as well as to uncover them in their use of AI. Codified approaches - top-down mandates written in a policy manual - cannot possibly embody the wisdom or insights of lived realities or keep up with rapid change. Cognitive psychologist and philosopher Mark Johnson has shown that ethics “conceived as a system of laws is simply too narrow and too unimaginative to captu