According to the Forbes Advisor survey, the most popular applications of artificial intelligence (AI) in business include customer service (56%), cybersecurity and fraud management (51%), customer relationship management (46%), digital personal assistants (47%), inventory management (40%) and content production (35%). Businesses also leverage AI for product recommendations (33%), accounting (30%), supply chain operations (30%), recruitment and talent sourcing (26%) and audience segmentation (24%).
To be clear, AI technology is relatively nascent. Sure, one variant, generative AI, has moved much faster than many others. But as Rita Sallam, distinguished VP analyst for Gartner said: "We’re at the early stages of a once-in-a-lifetime moment of disruption bigger than anything we’ve seen in our history, with the potential to reshape the competitive landscape, change work and level the playing field for every industry."
Declaring that generative AI has taken the business world by storm and making AI sexy once again, Kitman Cheung, APAC technical sales leader for IBM Technology Software, declared that in early 2023, references to AI on earnings calls were up 77% year-over-year and 75% of surveyed CEOs in a recent global study also believe the organisation with the most advanced generative AI will have a competitive advantage.
And AI solutions are no longer limited to the few big businesses with deep pockets. " AI is becoming increasingly accessible, with more dedicated solutions targeted at enhancing different business aspects," said Gavin Barfield, vice president & chief technology officer of solutions for Salesforce ASEAN.
Understanding the value of IT to the business
However, simply downloading a version of generative AI will not translate to harnessing the potential of the technology. Like other technologies, it requires a methodical approach to reap the business throughout its lifecycle.
As Cheung points out: "For AI models to be effective, useful, and trustworthy, they must be properly integrated into operational systems."
"How an organisation selects, governs, analyses, and applies data across the enterprise will define what they can do with AI. It’s not just about using data to improve AI, it can also help companies make better use of data and ensure the right data is chosen to power AI models."
Kitman Cheung
Barfield suggests companies start by analysing their processes, customer interactions, and data assets to pinpoint areas where AI can make the most impact. This will help to focus their efforts and optimise returns.
When to introduce AI in the workflow
That CFOs are interested in the technology is not being debated here. But as Mark D McDonald, a senior director analyst in the Gartner Finance practice observes – there remains scepticism over AI's impact on the finance function in general.
Editor's Choice: PodChats for FutureCFO: How finance should embrace AI
That scepticism is not limited to the finance function. In a poll of 126 members and candidates, (ISC)² revealed a consistently high degree of concern and scepticism about the increasing adoption and integration of AI and ML into all facets of consumer and business technology.
That AI projects fail falls to four reasons: unclear business objectives, poor data quality, lack of talent, and a lack of collaboration between and among teams.
Cheung says a broad range of offerings as well as the depth and breadth of customer engagement can make the process very complex.
"Organisations must first identify their own strengths—and areas where they could improve. AI should be introduced into the workflow when and where it can help increase organisation efficiency and drive revenue growth," he outlined.
Barfield suggests taking the time to assess the AI readiness of the organisation. This includes visualising specific use cases and desired outcomes, as well as addressing barriers to successful adoption.
"In so doing, companies will be able to make a more informed decision on the right moment to introduce AI and develop a realistic roadmap," he added.
Integrating AI without disrupting how work is done
Cheung recommends leaders outline a clear process for applying AI—starting with identifying the business problem the solution hopes to solve. "By setting clear goals for even experimental rollouts, companies can choose to advance only the most effective AI projects," he opined.
For his part, Barfield believes that a comprehensive proof of technology will determine whether the proposed AI solution will integrate into existing environments and demonstrate compatibility.
"For instance, companies can consider implementing AI tools in a controlled environment or start with pilot projects before a full-scale implementation," he suggested.
It's also a mindset
Asked how an organisation make AI a trusted technology among the people who will use it as a day-to-day tool, Cheung said companies must not let AI be an afterthought but instead infusing AI into the DNA of the enterprise, incentivising teams to look for ways to create business value with AI and igniting change at the grassroots level by developing a culture of continuous collaboration and a “use-case first” mindset.
Barfield adds that companies need to ensure that they are deploying tools that can meet the organisation’s data security and compliance needs.
"These tools should enable organisations to maintain data governance controls, improve accuracy through data grounding, mask sensitive data, ensure zero retention by the large language models (LLMs), check for bias and toxicity, and provide an audit trail."
Gavin Barfield
Scaling AI across the organisation
One observation FutureCIO and FutureCFO editors have noted in dialogues with leadership across Asia is the experience with bots in pursuit of robotic process automation: pilots are easy, scaling is hard.
According to recent Forrester data, only 52% of enterprises that have launched RPA initiatives have progressed beyond their first 10 bots. Principal analyst Leslie Joseph explains that the RPA market is becoming more varied, with smaller and more geographically diverse firms showing interest. While these firms understand the value RPA can bring, “they don’t necessarily want to go through the pain of figuring everything out on their own,” he points out.
Cheung acknowledges that while AI has today reached its inflexion point, organisations still struggle to successfully deploy responsible AI algorithms and models across real-world environments.
"90% of AI initiatives have yet to move beyond the test stages as companies struggle with scaling their AI across the enterprise. 31.6% of organisations said the biggest challenge is the technologies they are trying to use, both from a specifications and technical knowledge standpoint," he adds.
Cheung believes that for AI to drive truly impactful results across the business, it must integrate into existing workflows and systems, automating key processes across areas such as customer service, supply chain and cybersecurity.
"Enterprises need to be able to easily and securely move AI workloads around, and in today’s world that can mean across clouds, as well as modern and legacy software and hardware systems," he continued.
Barfield opines that scaling AI across the organisation requires leaders to clearly define AI goals and objectives, map them against wider business strategies and identify the most impactful use cases.
"This should be followed by assessing the organisation’s AI readiness and developing a clear AI strategy that plugs existing gaps. This includes priming the tech infrastructure, developing a data strategy that ensures quality data is used to train AI models, getting security and ethical guidelines right, building a culture of AI innovation and learning, and having a matching workforce development strategy," he elaborated.
Next step towards what AI means to the business
IBM's Cheung believes that for a business to realise the full potential of AI, it must be built on a foundation of trust and transparency, and it must be as widely available as possible, so all can benefit.
"Five pillars to trustworthy AI: explainability, fairness, robustness, transparency and privacy are the foundation for placing trust at every organisation’s AI roadmap," he continued.
For his part, Barfield says organisations can look to scale AI adoption by identifying additional areas where AI can boost efficiencies, deliver personalisation, improve collaboration and more. This must be supported by necessary infrastructure and resources – by expanding AI talent pools and equipping employees with the relevant soft and hard skills.
"They also need to find a trusted and secure way for employees to use AI. Salesforce has developed guidelines for the responsible development and implementation of generative AI that all organisations can adopt, with a focus on accuracy, safety, honesty, empowerment, and sustainability," he concluded.