Google was founded over two decades ago, but they released their first public set of ethical technology principles just last year. Facebook launched out of a Harvard dorm in 2004, but they formally launched an ethics program with a public investment last month. The era of tech companies moving fast and breaking things removed from public accountability is waning, if not entirely over. That’s precisely why it’s important for industry to understand–and admit in some cases–that there’s been a need for accountable, transparent, and companywide ethical practices in technology since the beginning.
The concept of ethical innovation is getting a great deal of public, government, and media attention. But a global conversation around how to design ethics for specific technology purposes has been going on for years. Experts continue to have a grand debates on this topic, but two important truths have emerged with time: 1) establishing corporate ethical frameworks and principles should not be an afterthought or reaction to public exposure; and 2) ethical tech practices should be designed by people who are building the products and solutions guided by independent, effective oversight.
Based on my experience, I believe there are four consistent elements that every company should take into account when designing ethical frameworks to govern innovation.
1. Select a committed leader
Companies need to select an internal champion to lead the effort, under the title of chief ethics officer. This person should be involved directly with building products or solutions on a daily basis. They should be seasoned, technical, and trustworthy. While CEOs and other C-suite executives should be fully invested in creating tech ethically, I would also argue that they should not serve as the champion. And the position definitely shouldn’t fall on anyone’s existing plate: Leading ethics initiatives should be a central part of the selected champion’s work, not a line item on their résumé or a corporate attempt to fill a void.
2. Build trust through transparency
Companies need to figure out how they will manage organizational design to safeguard employees and deliver transparency to stakeholders, investors, and customers. Companies that fail to deliver transparency tied to innovation will get caught if public issues arise with their technology. Internally, it’s crucial to develop policies and mechanisms for clearly communicating how, why, and when technologies will be introduced that change the status quo. Doing so will serve to ensure employees don’t feel surpassed by machines and everyone understands the complementary role(s) of automated technology.
3. Eliminate biases with testing
Companies must test and monitor technologies for biases, take stock of where they occur, remove data sets that create or perpetuate them, and put a human-led plan in action to correct them. The people, data, and technologies that power the global economy rely on human ingenuity to work. Bias is an unfortunate–and omnipresent–side effect of human behavior, mirrored in the machines we build. Companies that develop automated technologies and employ algorithms, in particular, need to operate with a sharp focus on preventing biases. In practice, ethics champions and their teams need to actively monitor data sets that inform algorithms and human-facing interactions, and take action to eliminate those biases as early as possible in the R&D cycle. Sometimes difficult, bottom line-connected decisions may need to be made, like halting development of a promising technology that continues to produce biased or offensive results.
4. Make ethics universal
Ethical innovation, and the frameworks that guide it, need to be focused, inspirational, and practical. People building technology should feel motivated to create ethical solutions from the start. And if they don’t, every person in the company should feel empowered to raise concerns to executive leadership when something feels off. Everyone, from the chief executive officer to the newest junior hire should be held accountable for operating within ethical frameworks. If a situation arises, for example, where the chief executive asks for something that falls outside of an established framework, the chief ethics officer should have the authority to deny the request on ethical grounds–and this should be broadly respected within the organization.
In order for ethical frameworks to translate into actual ethical innovation, companies need to determine the real impact of their tech on people and build internal mechanisms that correct or stem any negative side effects to their innovation. They need to create internal corporate structures that embrace accountability, transparency, and honesty. Designing ethical frameworks now will help companies around the world avoid issues that could impact the very survival of their business in the future. Ethical innovation needs to be the global norm–not a corporate afterthought.
Kriti Sharma is the vice president of AI and ethics at Sage.