Keeping humans in the loop when adopting AI

As Artificial Intelligence (AI) adoption continues to rise, it’s no surprise that the technology industry believes AI initiatives have delivered more value than expected. But at the same time, according to KPMG’s new report, Thriving in an AI World, 74 percent of technology business leaders still say that AI is more hype than reality, while 49 percent suggest adoption is moving faster than anticipated.

So why this disconnect?

Technology companies are at the forefront of creating technology platforms and building solutions using it. The importance of “Human in the loop” plays a huge role in the success and adoption of several of the AI solutions. In several instances, the benefits of AI have been over promised without understanding the broader governance needs around the solution. Take driverless cars as an example; the long term expectation is we will have a fully autonomous driving experience (Level 5), however the road to get there involves human engagement/intervention, which allows to collect and provide the required training data to improve the responses over time. There are several such examples in our everyday life, which emphasizes on reinforced learning based on “human in the loop” and broader governance for the adoption and success of these AI solutions.

As we think about the future, while 83 percent of technology business leaders reported AI is at least moderately to fully functional in their company, there are still hurdles to overcome:

  • Before you can do anything with AI, it’s critical to collect and organize your data, and have it machine ingestible (images, raw data etc.)
  • Many companies are still averse to moving to the cloud because of a variety of concerns, including privacy, but leveraging a hybrid cloud can address those concerns and yet drive benefits from these solutions.
  • Lastly, it is critical to think about innovation and the Digital journey. Begin with the problem statement, develop the Digital reference architecture and then look at tools and tech to address the problem statement. By developing an AI foundation, the building blocks for a solution and addressing a series of problem statements allows organizations to build a sustainable AI platform with reusable assets and lego blocks.

AI adoption will continue to progress, with the technology industry continuing to be at the forefront, but what’s most important to remember is that to increase adoption and solution, success will still need humans to realize its full potential.

To speak with Anand Sekhar about the study and AI in the technology sector,  please contact Melanie Batley.


About Thriving in an AI World

The KPMG study, Thriving in an AI World, is an evolution of a study KPMG originally released in early 2020. The findings are based on feedback from a range of 950 full-time business decision makers and/or IT decision makers* with at least a moderate amount of AI knowledge and at companies with over $1 billion in revenue**, across seven industries (150 respondents+ per industry): technology, financial services, industrial manufacturing, healthcare, life sciences, retail, and government. The online survey was fielded between January 3rd, 2021 and January 16th, 2021. The margin of error (MOE) for the total sample at the 95 percent confidence level is +/- 3.2 percentage points.

*Only government respondents included IT decision makers.

**Healthcare and life sciences respondents were from companies with over $100 million in revenue. +Healthcare and life sciences respondents only had 100 respondents per industry.




Anand Sekhar

Anand Sekhar

Managing Director, Digital Enablement, KPMG US

+1 312-665-1019

Related content