By my calculations, this might be the last entry in the Hey-Hi writing series. If the previous articles have been any pointer, it can be concluded that Artificial Intelligence (AI) has transformed industries around the world in ways we didn’t think were possible. And even scarier, its potential is far more than the impact it already has. AI has the potential to transform everything and allow more informed products and services, internal process efficiencies, healthcare, manufacturing, enhanced cybersecurity and reduced risk. However, to properly harness the impact of AI, and the extent to which it really does herald the creation of the fourth industrial revolution, it is necessary to address the regulatory and ethical challenges to its use.

If approached properly, AI can provide significant benefits for a firm, its customers and the world at large. Disruptive technology is a fact of life that has existed since the beginning of time. Widespread introduction of AI means we are beginning to see increased engagement from regulators with respect to AI, particularly in the financial services arena. In succeeding paragraphs, I will explore both the existing and developing regulations of AI, and key legal issues that arise for businesses deploying the technology that is available today. It is important to note that although this is a complex and evolving area requiring a multidisciplinary legal approach, none of it means firms should shy away from the use of Artificial Intelligence.

Many governments see great potential for AI to drive economic development and solve societal challenges and as such, they want to provide the legal framework needed to encourage innovation, attract investment and hence, enable growth. However, at the same time, they recognize there is a need to protect their citizens and address the ethical, legal, social and economic issues associated with AI. Given the specific risks AI poses, the challenge for law-makers is to understand both how existing legal and regulatory frameworks might apply. Currently, there is very limited specific AI legislation and as such, today’s use of AI is largely governed by the application of existing laws and regulation and to an extent, self-regulation by corporates by adhering, for example, to the voluntary ethical guidelines for AI published by Microsoft. Taking the use of AI in healthcare for instance, being relatively new, there are several unprecedented ethical concerns related to its practice. From data privacy to automation of jobs and representation biases, it is evident that as AI moves into areas of messier and less structured data, so will the challenges increase in magnitude and quantity. The major challenges include implementation and adoption of payment models, scaling these across healthcare systems, getting cultural institutions and clinicians to buy in and biggest of them all, figuring out where AI will sit in augmenting the intelligence of a clinician versus an independent diagnostic that runs on its own.

One other concern with AI includes recognizing that it has limitations. Given where AI is right now, it does not work very well where labels are not well defined, the quality of the training data isn’t good, or where the problem inherently is not solvable deterministically by the data the model is fed. I opine that we need to continue to maintain the humility to understand where it is going to work well, where it doesn’t, and how it complements what is already done well in each industry.

Debunking myths about AI

Fears and misconceptions about AI are heavily prevalent amongst a wide group of people. Worries range from preoccupations of job security as it relates to automation to science-fiction inspired fears of an impending, global robot insurgency. To placate these fears and dispel common myths about AI, it is vital that journalists are aware of the common misconceptions about AI and can counteract falsely held beliefs with facts. Some common misconceptions are:

• ‘Automation will put us all out of a job.’ Through history, transformations in the organization of labor have shown that human possess an innate ability to innovatively, and successfully, adapt to radical changes in the labor market. While automation undeniably poses a threat to certain contemporary livelihoods, overall gains in productivity and efficiency in the economy will present people with the opportunity to channel their creativity and expertise on forging jobs and career paths of their choosing.

• ‘Low-skilled workers will be replaced by AI and automation’ Adaption to the increased prevalence of AI in the workplace is, and will continue to take place across various sectors and industries, regardless of the workers’ socio-economic background and profession. While it is true that automation, facilitated through AI, has replaced humans in numerous low-skilled professions pursuing cost-cutting measures, the most profound implementation of AI has occurred in high-skilled professions. Taking the healthcare sector for instance, whether it be AI-assisted surgery, consulting medical applications or facilitating the management of records, AI’s emergence is assisting medical professionals (definitely not low-skilled) to carry out their jobs.

• ‘Artificial Intelligence will quickly overtake and outpace human intelligence’ A frequently peddled narrative amongst individuals skeptical about the development of AI involves the belief that robots and supercomputers will eventually reach a level of intelligence and performance that will surpass that of our own. I do not agree with this opinion because I believe a lack of sentience on the part of robots will preclude them from truly mastering the environment. Furthermore, the algorithms driving AI will continue to be vital in order to maximize gains from the use of AI.