Getty Images/iStockphoto

Why some businesses are failing at AI

Deloitte’s ongoing research into artificial intelligence adoption has found that more than 20% of organisations are underachieving in their AI ambitions

The latest edition of the Deloitte global study looking at artificial intelligence (AI) adoption in business has found that more organisations are struggling to achieve measurable value from the technology.

Deloitte’s State of AI 5th edition research surveyed global executives on how businesses and industries are deploying and scaling AI. The findings indicate that AI is moving ever closer to the centre of the enterprise. In fact, 94% of business leaders agree that AI is critical to success over the next five years. However, Deloitte’s research found that successful outcomes from AI seem to be lagging. 

The survey of 2,620 global business leaders between April and May 2022, reported that 79% of respondents said they achieved full-scale deployment for three or more types of AI application. But when asked to score the outcomes they attained against what their organisations had hoped to gain from the AI initiatives, 540 of the organisations surveyed achieved a score of only four or less, even though they had deployed five or more AI applications. From an AI strategy execution perspective, Deloitte classified these organisations as “underachievers”.

Looking at the low level of success among the underachievers, the authors of the Deloitte report recommended that organisations starting new AI projects should focus on business value. In fact, the research found that proving AI’s business value (37%) was the biggest challenge for all respondents.

Almost half of the business leaders surveyed said the main difficulty they faced was integrating AI into the organisation’s daily operations and workflows. As organisations attempt to scale up their AI projects over time, the Deloitte study reported that the key impediments business leaders faced were managing AI-related risks (50%), lack of executive buy-in (50%), and lack of maintenance or ongoing support (50%).

“This emphasises the resounding importance of clear leadership and focused investment that a successful AI transformation requires,” the report said. The findings are mirrored in the previous two years’ survey data.

While most respondents reported that they are happy with the payback period, Deloitte’s research found that those starting new AI projects have had varying success in proving business value, getting executive buy-in, ongoing support, and other factors needed for the longer-term, vast transformational opportunities that AI can offer. 

According to the report’s authors, the findings demonstrate the ongoing challenge of establishing the coordination and discipline needed to consistently fund initiatives after they have ceased to be the “shiny object”. 

Despite evidence that establishing clear processes and redefining roles to deliver quality AI result in improved outcomes, Deloitte found there has been little growth in the market in terms of adopting such practices. In both the fourth and fifth editions of the Deloitte study, just one-third of respondents reported that their companies are always following MLOps (machine learning operations), redesigning workflows and documenting AI model life cycles.

According to Deloitte, this lack of progress is significant because in both editions, operational leading practices were among the behaviours most highly associated with outcomes. The report’s authors said: “Leaders should embark on reinventing work to capitalise on the growing optimism and opportunity that their human workforce sees in AI. People are still at the core of a business’s success, and AI can help unleash the power of a combined human and machine workforce.”

Deloitte also recommended that for an organisation to become AI-fuelled, it needs discipline and focus to maintain systems and algorithms, so they can continue generating ongoing value instead of noise. This discipline and focus needs to extend to vigilant discovery and understanding of all associated challenges that may not be obvious in the early stages of an AI initiative. 

Read more about enterprise AI

  • Artificial intelligence is the fastest growth area in software. This is driving adoption, which will make AI mainstream technology in business software.
  • Enterprises looking to mature in their use of AI must focus on the information they’re putting into their models. Their models should create trust in their business.

Next Steps

How AI and marketing will influence our future

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management