There is increasing evidence of the business benefits of responsible AI (RAI), when companies mitigate risks through training and testing data, measuring model bias and accuracy, and model documentation. Companies that adopt responsible AI experience higher returns on their AI investment. Raj Shekhar writes that business leaders globally must coalesce around the imperative to develop rigorous, consistent standards for responsible AI adoption.
Much has been spoken and written about the risks to public trust and safety arising from the adoption of artificial intelligence (AI)-based applications across multiple sectors. In finance, the use of AI has led to discriminatory credit decisions. In healthcare, it has led to faulty medical diagnoses. In education, it has led to erroneous performance assessments. In agriculture, it has led to a widening gap between commercial and subsistence farmers.
This critical reportage has served to signal gaping holes in industry value chains and a lack of checks and balances that have been allowing subpar data and AI model governance practices within enterprises to continue. If humanity is truly to unlock the transformative potential of AI for a more prosperous, equitable, and sustainable future, the industry must adapt quickly to the escalating demands for building and deploying AI in a trustworthy, safe, and ethical manner.
In this article, I will attempt to explain why this is just good business. But before that, let us understand what we should mean by “good business” in this day and age.
Throughout recent history, the world has debated two competing economic models to define what a business ought to do for optimal wealth generation in a laissez-faire economy: shareholder capitalism and stakeholder capitalism. In the former model, serving the interests of shareholders through short-term profit maximisation is taken to be paramount by businesses. This model has led to tremendous growth and prosperity and has been instrumental in lifting billions of people out of poverty. Yet, it suffers from several major shortcomings that have resulted in inequalities of opportunity, income, and economic mobility, as well as social tensions, and environmental degradation to top it all.
Stakeholder capitalism is fundamentally a response to address the shortcomings of focusing exclusively on shareholders. In this model, the short-sighted practice of keeping shareholders happy at any cost is rejected to secure higher gains over the long-term in a socially beneficial and ecologically sustainable manner. But most importantly, no single stakeholder is given importance over others, and businesses are expected to optimise their economic performance by giving equal treatment to all stakeholders (including employees, suppliers, consumers, communities, shareholders, lenders, and regulators) and by accounting for the social and environmental impacts of their activities.
For quite some time now, multi-stakeholderism has been on the fast track to becoming the dominant face of capitalism that is socially responsible and sustainable by design and has been shaping business priorities globally for a more prosperous, healthier planet. The model gained widespread prominence and acceptance by global leaders when it became the key focus of the World Economic Forum’s 50th Annual Meeting in Davos in 2020. More recently, Larry Fink, chairman and chief executive officer of BlackRock, one of the world’s largest investment management corporations, in his 2022 open letter to CEOs and chairs of BlackRock portfolio companies, emphasised multi-stakeholderism as the indispensable value driver for businesses to ensure their long-term success. A “good business” today, therefore, is one that conducts its activities in adherence to the tenets of stakeholder capitalism.
It is this economic model that principally favours the prospect for enterprises to be able to sustainably derive long-term business value from investments in sophisticated responsible AI procedures, tools, and personnel that may appear costly in the short term. In other words, when building responsible AI strategies, enterprises just need to be mindful that no matter what the costs might be in the short term, the potential long-term benefits will most likely outweigh them.
In 2020, a survey report released by The Economist Intelligence Unit (EIU) had elaborately signalled this prospect. According to the EIU survey respondents, including senior industry executives and IT decision-makers in the United States, proactive interventions throughout the AI development lifecycle could help enterprises generate key value-adds in their production and innovation capacities and product and service offerings, lending them a strong competitive advantage. This, in turn, would enable enterprises to better cater to consumer values and expectations and demands for public trust and safety, build and retain shareholder and investor confidence, boost employee satisfaction and improve talent acquisition and retention, gain an upper hand in competitive bidding, and gather goodwill amongst regulators to serve as trusted industry advisors in consultations for future AI policies and potentially help prevent overly restrictive or regressive regulations.
These projections from the EIU report were so powerful that they began to come to pass just a year later when McKinsey, in its 2021 survey report on the global state of AI, reported higher returns from AI for companies that focused efforts on mitigating risks through training and testing data, measuring model bias and accuracy, and model documentation for responsible AI development and deployment. The business benefits from responsible AI are only becoming clearer by the day: very recently, in another global survey conducted by MIT Sloan Management Review, 70 per cent of surveyed business leaders reported that they were “already realising the business benefit of better products and services as a result of their organisation’s RAI [responsible AI] efforts”.
The need for enterprises to focus on the upside of scaling their responsible AI programs, therefore, cannot be overemphasised. Yet, the industry seems to be a long way off from adopting and implementing responsible AI practices satisfactorily. Overestimation of responsible AI maturity has been common amongst the majority of enterprises (55 per cent)—according to a global survey report released by Boston Consulting Group in 2021. Also, understanding of AI ethics issues amongst business heads, risk and compliance officers, and data and analytics staff at enterprises “remains patchy”, resulting in variable codes of responsible AI practice across companies—according to another global survey report released by FICO and Corinium Global Intelligence in the same year.
Business leaders globally must coalesce around the imperative to develop rigorous, consistent standards for responsible AI adoption by the industry. These standards ought to prioritise trust and ethics and oblige enterprises to integrate social and environmental impacts into their performance metrics by leaning on the precepts of stakeholder capitalism. This is just good business after all that will help us shape the “techade” for the benefit of humankind.