About 42% of enterprise-scale companies report having actively deployed artificial intelligence in their business and of those, 59% have accelerated their rollout or investments in the technology, according to IBM’s Global AI Adoption Index 2023. Enterprise-scale refers to companies with more than 1,000 employees.

“More accessible AI tools, the drive for automation of key processes, and increasing amounts of AI embedded into off-the-shelf business applications are top factors driving the expansion of AI at the enterprise level,” said Rob Thomas, senior vice president of IBM Software, in a statement. “We see organizations leveraging AI for use cases where I believe the technology can most quickly have a profound impact like IT automation, digital labor and customer care,” he said.

Even though the research revealed that 40% of companies surveyed remain “stuck in the sandbox,” Thomas said he is confident they will overcome barriers like the skills gap and data complexity this year.

What’s driving AI adoption

The top factors driving AI adoption are:

  • Advances in AI tools that make them more accessible (45%).
  • The need to reduce costs and automate key processes (42%).
  • The increasing amount of AI embedded into standard off-the-shelf business applications (37%).

Most surveyed companies (59%) actively deploying or exploring AI have accelerated their rollout or investments in the past 24 months. The top AI investments for organizations exploring or deploying AI are being made in research and development (44%) and reskilling/workforce development (39%).

For IT pros, the two most important enhancements to AI in recent years are tools that are easier to deploy (43%) and the increased prevalence of data, AI, and automation skills (42%).

Financial services is one of the most mature industries in AI adoption, followed by telecommunications, an IBM spokesperson told TechRepublic.

More stats from IBM’s Global AI Adoption Index 2023

The research also found that:

  • More than one-third of enterprise IT pros (38%) report their company is actively implementing generative AI and another 42% are exploring it.
  • Organizations in India (59%), China (50%), Singapore (53%) and the UAE (58%) are leading the way in the active use of AI, compared with lagging markets like Spain (28%), Australia (29%) and France (26%).
  • Companies within the financial services industry are most likely to be using AI, with about half of IT pros in that industry reporting their company has actively deployed AI. Within the telecommunications industry, 37% of IT pros state their company is deploying AI.

The main AI use case is automation

The AI use cases that are driving adoption for surveyed companies currently exploring or deploying AI cut across many key areas of business operations. Notably, automation is the main use case in several areas, including:

  • IT processes (33%).
  • Processing, understanding, and flow of documents (24%).
  • Customer or employee self-service answers and actions (23%).
  • Business processes (22%).
  • Network processes (22%).

Other areas where AI is being used include:

  • Security and threat detection (26%).
  • AI monitoring or governance (25%).
  • Business analytics or intelligence (24%).
  • Digital labor (22%).
  • Marketing and sales (22%).
  • Fraud detection (22%).
  • Search and knowledge discovery (21%).
  • Human resources and talent acquisition (19%).
  • Financial planning and analysis (18%).
  • Supply chain intelligence (18%).

The top barriers to AI use

Forty percent of companies surveyed  are exploring or experimenting with AI but have not deployed their models. The top barriers preventing deployment include limited AI skills and expertise (33%), too much data complexity (25%) and ethical concerns (23%), the company said.

Generative AI poses different barriers to entry from traditional AI models, the report noted. For example, IT pros at surveyed organizations not exploring or implementing generative AI reported that data privacy (57%) and trust and transparency (43%) concerns are the biggest inhibitors of generative AI. Another 35% also said a lack of skills for implementation is a big inhibitor, according to the report.

How to tackle AI’s barriers to entry

An IBM spokesperson said: “Companies need to set AI strategies that clearly define the problems they want to solve, make sure they have the correct data in the right place to drive those outcomes, overcome skills gaps by selecting the right people and automation tools, and incorporate AI governance from the start of their adoption process.”

For some organizations, the best approach might be starting small and targeted. “In 2024, we expect business leaders to begin analyzing and testing AI on a case-by-case basis and not paint a broad stroke,” the spokesperson said, “assuming the technology is the right tool to solve every problem. Businesses using AI for the first time will use off-the-shelf AI assistants built for specific business needs.”

AI’s impact on the workforce

Among surveyed organizations, one in five reported they do not have employees with the right skills to use new AI or automation tools, and 16% cannot find new hires with the skills to address that gap.

Companies using AI to address labor or skills shortages said they are tapping AI to reduce manual or repetitive tasks with automation tools (55%) or automate customer self-service answers and actions (47%). Only 34% said they are training or reskilling employees to work together with new automation and AI tools.

The importance of trustworthy and governed AI

According to IBM’s report, IT pros understand the need for trustworthy and governed AI, but the aforementioned barriers are making it difficult for respondent companies to put into practice.

For example, the research found that IT pros agree overall that consumers are more likely to choose services from companies with transparent and ethical AI practices (85% strongly or somewhat agree). They said the ability to explain how their AI reached a decision is important to their business (83% among companies exploring or deploying AI).

However, a startling finding was that — even as many companies already deploying AI are facing multiple barriers in the process — well under half reported they are taking key steps toward trustworthy AI, such as:

  • Reducing bias (27%).
  • Tracking data provenance (37%).
  • Making sure they can explain the decisions of their AI models (41%).
  • Developing ethical AI policies (44%).

Reducing bias starts with governance, the IBM spokesperson said. “To harness AI’s full potential and reduce bias, data and AI governance tools are essential to scale models while maintaining fairness, transparency and compliance,” the spokesperson said.

“Without these safeguards, AI outputs can be biased, discriminatory — or sometimes just plain wrong,” the spokesperson added. Without the use of governance tools, AI can expose companies to several data privacy issues, including leaking proprietary data and sensitive data or infringing on copyrights. Organizations also run the risk of legal complications and ethical dilemmas, so incorporating governance from the start can help avoid problems later.

IBM’s survey methodology

IBM said the survey was conducted in November 2023 among a representative sample of 8,584 IT professionals in Australia, Canada, China, France, Germany, India, Italy, Japan, Singapore, South Korea, Spain, UAE, U.K., U.S. and LATAM.

Subscribe to the Innovation Insider Newsletter

Catch up on the latest tech innovations that are changing the world, including IoT, 5G, the latest about phones, security, smart cities, AI, robotics, and more. Delivered Tuesdays and Fridays

Subscribe to the Innovation Insider Newsletter

Catch up on the latest tech innovations that are changing the world, including IoT, 5G, the latest about phones, security, smart cities, AI, robotics, and more. Delivered Tuesdays and Fridays