AI is in the driving seat but firms failing to visit ethics
Business chiefs may have high expectations for artificial intelligence but only half of companies have policies in place to identify and address the ethical considerations of the technology.
So says a new Cognizant report, "Making AI Responsible - and Effective", which analyses the responses of almost 1,000 executives across the financial services, technology, healthcare, retail, manufacturing, insurance and media and entertainment industries in Europe and the US.
The study shows that nearly two-thirds (63%) of business leaders believe that AI is extremely or very important to their companies today, and 84% expect this will be the case in three years' time. Lower costs, increased revenues and the ability to introduce new products or services, or to diversify, were cited as the key advantages for the future.
Companies that are growing faster than the average business in their sector are particularly bullish, with 86% of executives of these companies stating AI is extremely or very important to their success, compared with 57% of those at their competitors with slower growth.
These industry leaders say they plan to use AI to drive further growth, solidifying their leading positions and pulling even further away from the pack. This is reflected in their greater investment in key AI technologies, including computer vision (64% vs 47%), smart robotics/autonomous vehicles (63% vs 43%) and analysis of natural language (67% vs 42%).
Almost half of those companies (44%) undertaking AI projects expect to increase their staff headcount over the next three years, with retail and financial services industry executives most likely to expect a boost to employment (56% and 49% respectively).
However, while many executives talk a good game, of the two-thirds who said they knew about an AI project at their company, only 24% – just 15% of all respondents – were aware of projects that were fully implemented.
Some 40% of respondents said that securing senior management commitment, buy-in by the business and even adequate budget remain extremely or very challenging, indicating that many companies are not yet fully committed to AI's central role in advancing business objectives.
Meanwhile, only half of companies have policies and procedures in place to identify and address ethical considerations – either in the initial design of AI applications or in their behaviour after the system is launched.
The report provides three key recommendations to help companies take action and achieve the significant business benefits of AI:
Formulate AI strategies. These should focus on opportunities that promise measurable value – not only reduced costs and increased revenue, but also benefits such as improved customer service, entry to new lines of business and enhanced employee experiences. It is especially critical that strategies take a human-centric view of AI so that machines can work successfully alongside and for people.
Develop governance structures. Companies will need to work proactively to ensure that AI decision-making is transparent to those involved, that AI earns trust by avoiding errors and data-driven biases; and that AI is personalised and able to provide tailored, relevant and context-aware support as it interacts with humans.
Create and maintain responsible AI applications. Because of AI's potential power, ethical concerns need to be interwoven into everything companies do with the technology. That means building AI systems ethically, and monitoring to ensure that those systems operate ethically over time, even as the AI applications learn and evolve. To be successful, companies will need to boost their ethics-related efforts upfront as AI touches more and more parts of business and society.
Cognizant digital business European head Sanjiv Gossain said: "The challenge today is less about understanding technical questions and technology capabilities, and more about crafting a strategy, determining the governance structures and practices needed for responsible AI."
He insists that companies need to pay more attention to the non-technical considerations of AI deployments, many of which are more critical and complex than those related to developing and running the technology itself.
Gossain added: "AI operates in the real world and this will not only help companies determine which technologies can be used to advance business objectives, but also those which have the potential to irritate customers, alienate employees, drive up R&D and deployment costs, and undermine brand reputation."