Companies are rushing to adopt the growing number of artificial intelligence (AI) tools available, but most aren’t implementing the metrics needed to measure return on investment.
IBM research shows that many businesses lack a comprehensive AI strategy and buy products primarily based on features. AI Readiness Barometer Survey The report, released this week, found that only 17% of companies have a clearly defined AI strategy, while the majority, 38%, are still in the process of developing one. Another 30% have an AI strategy focused on specific use cases, while 7% ultimately abandoned their AI strategy or admitted they were unable to implement it effectively.
According to the report, the spread of AI-enabled business applications has led to approximately 43% of companies adopting AI. Ecosystemconducted a survey of 372 technology and business leaders across five ASEAN markets: Singapore, Indonesia, Thailand, Malaysia and the Philippines.
Also: IT leaders are concerned that rushing to adopt the AI generation will have a negative impact on technology infrastructure.
Additionally, while 85% acknowledge the power of AI, only 22% measure its value and pay attention to reporting on it, meaning most businesses don’t have clear ROI (return on investment) metrics to determine whether their AI investments translate into internal efficiencies or increased external revenue.
And there’s a gap between how organizations assess their AI readiness and the reality of this situation as assessed in the survey, Ecosystem CEO Ulrich Loeffler said at a press conference in Singapore. He explained that the research firm collected data to assess an organization’s readiness and maturity in deploying its AI roadmap across four criteria. Criteria include culture and leadership, data foundation, and governance framework. The scores were tallied and used to categorize organizations into one of five AI readiness stages: “Traditional,” “Emerging,” “Integrated,” “Transformative,” and “AI-First.”
While 39% of respondents described their companies as in the Transformation stage, Ecosystm’s assessment placed only 4% in this category. Another 16% of companies said they were AI-first, but Ecosystm concluded that only 1% fell into this stage of AI readiness.
Also: It’s time for businesses to get beyond the hype of generative AI and discover true value
AI-first organizations rank highly in four key areas, including governance, have a dedicated role overseeing the function, and develop ethical AI solutions. These companies also have a data-centric strategy that provides seamless data access and an AI-powered workforce that includes centralized data teams with strong AI and machine learning capabilities.
In explaining why so few companies have made progress with AI adoption, Loeffler noted that while it’s easy to achieve a proof of concept, it can be difficult for companies to scale their AI adoption.
Additionally, he highlighted the need for organizations to monitor and evaluate the impact of adoption to ensure AI applications are delivering the intended benefits.
According to the survey, 63% of companies are using AI to power intelligent document processing, 60% are using the technology for support and helpdesk applications, 57% are using AI for payment and billing automation, a further 56% are using AI for technology documentation, 55% are using AI for content strategy and creation, and 55% are using AI for recruiting purposes.
Nearly 25% of organizations note that identifying use cases to run pilots or proofs of concept is their top AI priority, 22% see improving data quality, interoperability and consistency as an AI priority, and 21% say they need to upskill and reskill their workforce to be data-aware.
About 39% of respondents said their companies have limited AI expertise and few subject matter experts, while 26% said they use AI within existing applications or platforms and do not have standalone AI capabilities.
Also: Amid growing interest in AI, there is a significant lack of transparency
Additionally, the survey highlighted the lack of a governance framework as a concern, with only 18% of organizations having a dedicated role in AI and data governance, while 66% of organizations have this responsibility spread across departments or teams, and around 3% have no clear policies or responsibilities defined for AI governance.
Additionally, the report found that only 12% have a process in place to track fluctuations in AI model performance or model drift that can affect outcomes over time.
“The tangible benefit for organizations is in scaling AI to accelerate innovation and productivity,” said Katherine Liang, general manager, ASEAN, IBM. “Unfortunately, many technology and business leaders overestimate their organization’s ability to successfully adopt AI. Addressing AI requires strong leadership, a robust data strategy, the right people to execute it, and a well-thought-out governance framework to ensure responsible and ethical use of AI.”
“Without this strong foundation, organisations risk making implementations that focus solely on the capabilities of the technology without considering the long-term impact to the business,” Liang said.
Also: AI risks are everywhere, and now MIT is adding them all to one database.
Hans Dekkers, general manager of IBM’s Asia Pacific region, also pointed out that organizations need AI, along with automation, to keep up with the speed of change.
ZDNET asked whether the risk of incidents like the CrowdStrike outage is increasing, and whether organizations should rely more on automation to handle patch management and other critical work processes.
Dekkers said automation is crucial for freeing employees from time-consuming, repetitive tasks and speeding up the deal process.
Also: The race for global AI leadership may come down to governance
But automation needs to be implemented correctly to avoid failure, he said.
Loeffler added that this should be part of an organization’s governance framework, including ensuring that third-party AI applications meet the company’s AI safety policies.
