Presented by Hypergiant
Every major company and government deploying AI at scale has publicly declared some form of statement on ethics in AI and articulated a set of guiding principles. In fact, Harvard’s Berkman Center for Internet & Society has launched a project to catalog all of these statements.
While statements of ethics in AI are public, how ethics in AI is being implemented is still as black box as the algorithms and tools being created. The AI Now Institute at NYU, has astutely pointed out in its 2019 report “…the vast majority of these (statements of ethics in AI) say very little about implementation, accountability, or how such ethics would be measured and enforced in practice.”
Transparency in execution of ethics in AI is vital to stakeholder trust in the growth and development of AI and to an organization’s competitive positioning. In fact, the World Economic Forum will be meeting in Davos at the end of the month to explore this very issue. They have declared ethical standards to be as important as technical standards in creating and defending competitive advantages in an AI-driven world.
No doubt, in 2020 it’s imperative that corporations start implementing an ethical framework for AI. You can use these three steps to create a competitive advantage:
1. Align AI uses cases to overall vision and values
The first, and most important, step in implementing ethics in AI is aligning AI use cases with the overall vision and values of your organization. If you are fortunate, your organization is already good at communicating and putting its vision and values into practice. It all starts in the C-Suite.
At Hypergiant, our Co-Founder and CEO, Ben Lamm, has clearly defined our over-arching vision as, “Delivering on the future we were promised.” This over-arching vision is governed by a clearly enunciated set of values and guiding principles that inform everything that we do. We call this approach Top of Mind Ethics (TOME). Similar to back of the envelope financials and the classic elevator pitch, top of mind ethics should be simple enough to use that any relevant stakeholder can apply it to each AI case. This is only the first step and should be buttressed with a clear and coherent ethical decision-making process that aligns each AI use case to an organization’s vision and values.
2. Apply an ethics framework to every AI use case
All of our AI uses cases at Hypergiant are vetted through Immanuel Kant’s deontological framework. The core elements of this framework are: Goodwill. Does the AI use case have a positive intent? Does it meet the Categorical Imperative? If every company in our industry and every industry in the world deployed AI in this way, what would the world look like? And finally, does it pass the Law of Humanity? Are people being used as a means to an end? Or are people the primary beneficiary of this AI use case?
Once the use case owner makes an ethical case for its project, we utilize a cross-functional Red Team to raise objections to each step of the ethical decision-making workflow and the burden of proof falls on the use case owner to satisfactorily answer any issues raised and design its use case to address any objections. Our Red Team is comprised of some of the sharpest people in our company from strategy, design, data science, engineering, sales, and legal. These arguments are then submitted to an Ethics Review Board for final sign-off.
Our engineering and design teams have developed a tool to guide this ethical reasoning workflow and it allows us to have a record and archive of all of our AI use cases and the ethical decision-making that goes into vetting our projects. This tool is also a reference library for use case owners to query and model as they develop future use cases for ourselves and clients.
3. Evangelize and iterate on the ethical-decision-making process
The ability of organizations to implement, evangelize, and iterate their ethical decision-making will determine its competitive position in an AI-driven world. Harvard Business Review spells it out pretty succinctly. “The potential for businesses that embrace digital operating models is huge, but the capacity to inflict widespread harm needs to be explicitly considered. Navigating these opportunities and threats will be a real test of leadership for both businesses and public institutions.”
Unlike company trade secrets such as algorithms and code, ethical-decision-making should be transparent to all stakeholders, including society-at-large. When the C-Suite embraces ethics in AI, it makes it clear to all members of an organization and its customers that ethics is a central part of the culture and will be reflected in its workflow. Evangelizing this process will increase trust in the end products and solutions deployed. It is not an exaggeration to say the future of the organization and its impact on the world depend on it.
Will Griffin is VP of Ethics & Diversity in AI at Hypergiant Industries.
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. Content produced by our editorial team is never influenced by advertisers or sponsors in any way. For more information, contact firstname.lastname@example.org.