How is AI adoption playing out around the world, and how do adopters and governments view the technologies’ potential—and risks? Recent research from Deloitte Global member firms shows executives facing a wide range of challenges.
Artificial intelligence has evolved from an esoteric research topic—with its origins six decades ago in corporate and academic computer science labs—into a collection of powerful technologies with mainstream business promise and applicability. Deloitte’s global AI study finds that, in organizations adopting AI, more than eight in 10 leaders see AI as “very” or “critically” important to their business success in the next two years.1 AI adoption and spending are surging globally. According to one report, 37 percent of organizations have now deployed AI—a 270 percent increase from four years ago.2 Analysts project global spending on AI to top US$35 billion in 2019 and more than double to US$79.2 billion by 2022.3
Explore the AI & cognitive technologies collection
Subscribe to receive related content from Deloitte Insights
Download the Deloitte Insights and Dow Jones app
What is driving this tremendous upswing? Many foresee AI helping to spur enormous productivity gains over the next decade, making it essential to the competitiveness of national economies.4 Some researchers even believe that AI is poised to become a “general-purpose technology”—one of only a couple dozen inventions in human history (including steam engines and the internet) to have pervasive effects across industries, spark complementary innovations, affect economies, and actually change societies.5 Technology giants with global reach are declaring “AI-first” ambitions and using AI innovations to reduce costs, increase productivity, and fuel new products and services. At the same time, nimble, AI-focused startups pose a competitive threat to traditional businesses.
Companies of all kinds across the globe increasingly regard the development of strong AI capabilities as essential to remaining competitive. National governments, too, are focusing on the economic potential of AI for their countries. Of course, AI technology adoption won’t necessarily materialize effortlessly into economic gains all around, and to fully capitalize on AI’s promise, leaders should strive to get execution and strategy right. They should first aim to understand the best use cases for their industry and particular circumstances, and then build from there. Leaders should also be aware of possible risks associated with AI and develop plans to manage them.
Recent research from Deloitte global member firms around the world—which we describe and excerpt in this report—looks at AI from a variety of perspectives. How is AI adoption playing out around the world, and how are adopters and governments viewing the potential—and risks—of AI technologies? What kinds of business benefits are adopters realizing? Are they using AI to keep up with the competition or to pull ahead? Are they using it to evolve their businesses—improving processes and products—or to transform their companies (and even industries) with innovative solutions? What are the challenges and risks that organizations confront as they create business value with AI?
Given AI’s tremendous potential to drive economic expansion and alter the nature of work, it’s unsurprising that many nations are engaging in a new kind of space race to become AI leaders and reap the spoils. Twenty-six countries (and counting) have published national AI strategies or frameworks to foster growth, and many are backing up their ambitions by making investments, setting up programs, sponsoring research, and establishing partnerships.6 Many governments are also assessing how they can ensure privacy, safety, transparency, accountability, and control of AI-enabled systems without stifling innovation and the potential economic benefits.
A recent Deloitte report, Future in the balance? How countries are pursuing an AI advantage, provides a global view of AI early adopters, based on surveying 1,900 executive respondents from seven countries. These adopters are ramping up spending on AI technologies and reaping positive returns. They profess a growing belief that AI will be critically important to their business success—indeed, almost two-thirds of the executives report that AI technologies enable their companies to establish a lead over the competition.7
Globally, three-quarters of respondents agree that human workers and AI technologies will augment each other to produce new ways of working, and that AI empowers their employees to make better decisions. And augmenting and amplifying human intelligence is likely to be only the beginning: According to Deloitte Consulting’s AI-fueled organizations: Reaching AI’s full potential in the enterprise, AI is progressing toward autonomous intelligence, in which processes are digitized and automated to a level where machines and systems can act directly upon the intelligence derived from them, without human involvement.8
Remarkably, a majority of early adopters within each country believe that AI will substantially transform their business within the next three years. However, as pointed out in Is the window for AI competitive advantage closing for early adopters?—part of Deloitte’s Thinking Fast series of quick insights9—the early adopters also believe that the transformation of their industry is following close on the heels of their own AI-powered business transformation.10 Globally, there’s a sense of urgency among adopters that now is the time to capitalize on AI, before the window for competitive advantage closes.
However, comparing AI adopters across countries reveals notable differences in AI maturity levels and urgency. While many nations regard AI as crucial to their future competitiveness, these comparisons indicate that some countries are adopting AI aggressively, while others are proceeding with considerable caution—and may be at risk of being left behind.
Consider Canada. At the national level, Canada is making efforts to boost the nation’s AI prowess: In 2017, the government initiated its C$125 million Pan-Canadian Artificial Intelligence Strategy to help drive research, talent development, and innovation.11 Canada is home to world-class AI research institutes and thriving AI startups in cities such as Montreal and Toronto.12 More broadly, though, at the company level, early AI adopters display a lack of urgency: Deloitte’s global study found that only 5 percent of AI adopters in Canada rate AI as “critically important” to their business today, and 27 percent expect it will be in two years—the lowest level among countries studied.13 Concerns around AI risks may be putting a damper on leaders’ efforts; for example, 48 percent of early adopters from Canada cited “making the wrong strategic choice based on AI recommendations” as a top-three AI risk—highest among countries studied.
The recent Deloitte Canada report Canada’s AI imperative: From predictions to prosperity highlights the importance of AI to that country’s economic future and exhorts businesses to get off the sidelines. Why is there so much at stake even for a comparatively small economy such as Canada’s? Because the country, the authors explain, has an opportunity to help shape the economic and social conditions surrounding AI and its applications. Canadian companies can either jump in now and help create the AI future—or wait for others to design that future to their own advantage. The report urges the nation not only to capitalize on its research, skill, and startup strengths but to begin establishing a world-class AI ecosystem.14
At the other end of the spectrum, China has a grand strategy to help build the AI future. Boldly declaring China’s ambition to become the world’s leading AI innovator by 2030, the Chinese government has published a national AI strategy and announced plans to invest tens of billions of dollars in AI research and development.15 In 2017, nearly half of global AI venture funding went to Chinese AI startups, outpacing US startups for the first time.16 China reportedly has the world’s second-highest number of AI companies, nipping at the heels of the United States.17 According to the Deloitte global AI study, 54 percent of AI early adopters in China consider AI to be “very” or “critically” important to their business success, and that is expected to rise to 85 percent in two years—the highest level across countries studied. They are also more likely to believe AI is helping them open a sizable edge over their competition (55 percent, versus 37 percent globally).18
China faces speed bumps along its journey to AI leadership: As Deloitte’s 2019 TMT Prediction China inside: Chinese semiconductors will power artificial intelligence explains, China is the world’s leading consumer of semiconductors, but its domestic manufacturers can meet only a small portion of the nation’s demand.19 Amid uncertainty brought on by international tension and worry about supply chain interruptions—and the increasing demand for specialized chips to run AI algorithms—China is making development of its own chip industry a strategic priority.20 To help increase technological self-reliance and close the gap between Chinese manufacturers and global chip leaders, the Chinese government has established a state-backed semiconductor-focused fund, the “Big Fund,” which reportedly raised US$29 billion in its second financing round in 2018, following an initial US$21.8 billion round in 2014.21 Because chip industry development is a complex, deliberate process, some analysts caution that China’s returns on AI investment may be lower and slower in the future.22
Acting decisively on AI now, as some countries and early-adopter companies are, could be critical for future competitiveness. Leaders should aim to do so mindfully, balancing action with a suitable amount of caution. To capitalize on AI while also avoiding stumbling blocks, they should get two areas right: mastering execution and managing AI risks.
To succeed with AI, getting execution right from a technology and organizational standpoint is critical, and mastery may take time and experience. Successful adoption requires pursuing the right use cases, developing a strong data foundation, and forging an AI strategy. Early adopters confront a set of challenges as they attempt to create business value with AI, including implementation and integration difficulties, data challenges, and substantial shortages of AI skills. Indeed, more than two-thirds of executives who responded to the Deloitte global AI survey report moderate-to-extreme AI skill gaps.23
Deloitte analysis has identified a group of companies that are having the most success with AI—and can provide guiding principles for other AI adopters to follow. Our report, Seasoned explorers: How experienced TMT organizations are navigating AI, presents a deep dive into US-based AI adopters from technology, media and entertainment, and telecommunications (TMT) companies.24 Compared with their counterparts in other industries, TMT companies are spending significantly on AI technologies—and getting higher returns. Moreover, our analysis finds striking differences between a leading group of more mature AI adopters in the TMT industry—those that have implemented more AI systems and developed a high level of expertise—and those with less experience. Facing substantial job changes and a severe AI skills shortage, the “seasoned” adopters are vigorously training and educating their current workforce to succeed with AI, and seeking software developers, business leaders who can interpret AI results, and change management experts.
Curiously, the least experienced adopters, perhaps feeling that they need specialists to build every AI system from the ground up, are most eagerly searching for heavy-duty AI researchers. This is not unlike tasking oneself with designing and building an airplane before one can fly. Leaders should consider speed and ease of adoption.
With the window for competitive advantage closing quickly, not everyone has time or resources to build AI systems from scratch. Deloitte’s 2019 TMT Prediction Artificial intelligence: From expert-only to everywhere explains the growing popularity of AI-infused enterprise software and cloud-based AI services, both of which provide a way to develop and scale AI projects quickly, even without a strong IT infrastructure, extensive AI and data science expertise, or deep pockets.25 Large technology companies that have been at the forefront of the AI revolution are now competing with each other to make AI easier to use. In fact, cloud-based AI is beginning to democratize AI technologies, making capabilities and benefits available even to less-well-funded businesses and less-experienced adopters. This is paving the way for more widespread AI adoption. Less-experienced adopters and those still sitting on the sidelines should take note.
While offering the potential for great reward, emerging technologies also necessitate understanding and managing risks—and AI is no exception. Deloitte’s US-based report State of AI in the Enterprise, 2nd edition: Early adopters combine bullish enthusiasm with strategic investments examines potential AI risks that concern executives, including cybersecurity, legal and regulatory, and ethical risks.26
There’s a growing awareness of the diverse ethical risks that AI may pose. Among the AI adopters surveyed, the top ethical worry is the power of AI technologies to help create or spread false information. A report from Deloitte Germany, Cognitive artificial intelligence: The invisible invasion of the media business, explains how AI technologies can themselves be used to combat textual “fake news” and violent or offensive images and videos faster and better than humans can.27 Unfortunately, as Deepfakes and AI: Questioning artificial intelligence ethics and the dangers of AI explains, the power of AI can also be misused to create highly realistic fake images, audio clips, and videos—for example, convincingly manipulating authentic video to “put new words in someone’s mouth.”28 Identifying deepfakes is currently difficult for both humans and AI, and although detection methods continue to improve, some AI researchers admit they’re currently “outgunned” in the battle.29 With the potential for deepfakes to affect elections, national security, financial markets, and even international relations, the stakes are enormous.
Another issue becoming an ethical concern for executives is the lack of explanation or audit trail for some AI-powered decision-making. Deloitte Netherlands takes a deep dive into this issue in A call for transparency and responsibility in artificial intelligence, advocating for “transparent AI.”30 Why is it so crucial that AI results be “explainable”? For one, it allows humans to understand (and explain to others) why specific decisions have been made by AI and to assess whether they make sense. Second, explainable AI enables both technical staff and executives to understand how AI systems—which may be obtained externally, such as via open source or as a cloud service—arrive at decisions, therefore mitigating risk to the company.
Other ethical risks that concern AI adopters include unintended consequences of AI decisions, misuse of personal data, and potential bias. A recent report from Deloitte US, Can AI be ethical? Why enterprises shouldn’t wait for AI regulation, explores this thorny ethical landscape and calls for urgency in designing approaches and mechanisms to address these risks. The report also reviews how technology vendors, corporations, academic institutions, and governments are already laying the groundwork for ethical AI use.31 AI ethics: The next big thing in government, a perspective from Deloitte Middle East, makes the case that ethics efforts need to transcend national borders. Because of AI’s potential to affect billions of people globally, the authors advocate for a global AI ethics framework: While acknowledging the complexity of developing a code of ethics that’s accepted globally, they believe responsible AI development and deployment will require an international regulatory model that smartly secures AI technologies’ benefits for societies and economies.32
Finally, AI ethics: A new imperative for businesses, boards and C-suites, a report from the Deloitte Risk & Financial Advisory practice and the Notre Dame Deloitte Center for Ethical Leadership, puts forward the point of view that everyone involved in advancing AI—from corporate boards and management, to researchers and engineers—shares responsibility for applying ethical constructs throughout the AI product life cycle. To that end, the authors offer a framework outlining four dimensions of ethical concern that leaders may consider as their organizations design and build AI systems.33
AI has already begun to transform industries. In health care, providers are using AI technologies to review radiology scans and pathology samples more accurately, accelerate drug discovery, identify who may be at risk of developing a condition, and even detect diseases at earlier stages. In financial services, firms are using AI both in the back office (to reduce risk in credit underwriting and to detect possible fraud) and in the front office (for conversational banking and personalized customer experiences). In retail, AI technologies are transforming shopping via virtual assistants and customized recommendations. Manufacturers are using AI to optimize products and processes and detect potential problems before they occur, improving performance and uptime.
Several reports offer deep dives into the use of AI in various industries. The Deloitte Digital publication From mystery to mastery: Unlocking the business value of artificial intelligence in the insurance industry examines the use of AI in insurance, concluding that insurance is lagging behind other industries. The industry’s recent focus has been on using AI to automate repetitive tasks, thereby contributing to improved operations and customer efficacy. However, the abundance of data in insurance holds great promise for insurers to use AI as a competitive differentiator, and the authors provide an extensive list of potential future use cases for AI technologies across the insurance value chain.34
The new physics of financial services: How artificial intelligence is transforming the financial ecosystem, based on more than 200 interviews with global experts, describes the great upheaval that AI is causing in that industry—affecting operating models, disrupting competitive dynamics and strategies, and raising public policy issues. Whereas economies of scale and standardized products used to give financial institutions an edge, in the AI era, highly customized products and interactions can drive revenue. Financial institutions should adapt to new ways of acquiring and retaining customers, as well as a changing competitive landscape.35
Though government agencies are adopting AI less rapidly than private sector organizations, the public sector is employing AI technologies for use cases such as virtual assistants and airline flight tracking. According to the report AI-augmented government: Using cognitive technologies to redesign public sector work, AI is poised to transform not only how government employees do their work but the very nature of that work. According to the authors’ analysis, the US federal government could free up hundreds of millions of working hours annually by automating public sector tasks.36
Finally, Intelligent IoT: Bringing the power of AI to the Internet of Things explores the growing importance of AI to the Internet of Things (IoT), which has applicability in a wide range of industries, including manufacturing, transportation, automotive, utilities, government (“smart cities” initiatives), and even health care.37 Machine learning, an AI technology, is being used to identify patterns and aberrations in the data generated by IoT sensors and devices, and can make predictions faster than was previously possible. Companies are also using other AI technologies such as computer vision and speech recognition to rapidly derive insights from IoT data. AI is thus acting as a “force multiplier” for IoT products and deployments, helping companies make their operations more efficient, create new products and services, avoid unplanned downtime, and improve risk management.
The following excerpts from Deloitte Global member firms’ recent research provide further insights into the state of AI in the enterprise around the globe. Taken together, these reports suggest activities and strategies for organizations to consider as they aim to fully capitalize on the promise of AI.
Research from Deloitte global member firms suggests that AI holds enormous promise to transform businesses and entire industries. We’ve seen that many countries and companies believe AI is essential to their future competitiveness. With AI use on the rise globally, early adopters are already feeling that their industry competitors are starting to catch up, and with easier, cloud-based ways to develop AI solutions growing in popularity, the competitive pressure will likely keep building.
On an increasingly crowded AI playing field, which adopters will be most successful in capitalizing on the promise of AI and maintaining a competitive edge? We see some signs in our research. Organizations should strive to:
Ultimately, success will likely depend on each organization determining how AI can improve its operations, the way humans and machines collaborate, and what it sells. It’s not just companies that should formulate strategies for using AI. Nations should decide their AI approach as well, lest they find themselves in a future designed and created by others.
Will AI transform industries and form the basis for economic competitiveness in the future? Nothing is certain, but many countries and companies are building AI into their future.