IT Brief Australia - Technology news for CIOs & IT decision-makers
Story image
Caution needed in AI implementation for ANZ businesses
Wed, 10th Jul 2019
FYI, this story is more than a year old

Has your IT team and customer service division jumped on the AI bandwagon yet?

Or are you concerned that not having already done so has put you behind the eight ball?

It's an understandable concern, given the efficiency gains and productivity enhancements AI-powered solutions promise to deliver.

Chief among them is an immersive user experience that can develop as a result of understanding customer behaviour better and capturing the right feedback from time to time.

But racing headlong into widespread adoption of AI-driven solutions can be a trap for the unwary. The fact is moving slowly and prudently towards the adoption of AI solutions is a strategy that's likely to serve an organisation better than a fast-paced embrace of AI-centric architecture.

Some of the world's biggest tech players have learnt the perils of letting AI roam freely within their enterprises.

They were later forced to put the muzzle on their AI-powered bots very quickly after the systems began responding to their customers and social media posts with its own offensive and racist statements.

How an equivalent scenario would play out in an enterprise is a question well worth considering before implementing AI, not after, as it could behave in a way that doesn't fit with an organisation's corporate culture and ethos.

 AI everywhere

IT consultancy company Deloitte has named 2019 the year in which AI will move beyond the purview of the experts and become embedded in mainstream solutions and business practices within Australian organisations.

Natural language processing and conversational AI are expected to take a firm foothold in the contact centre management sphere.

"As the adoption of AI from 'expert only' to 'everywhere' increases, organisations must focus on good governance and the responsible use of AI," Deloitte's 2019 predictions report states.

It starts with good data and the right tools

Given the malleability of AI, the onus is on businesses to ensure they manage and consistently modulate their newly-implemented AI solutions.

Doing so will ensure they end up with an unbiased and responsible asset—the digital equivalent of a sober, industrious employee, rather than a thoughtless, out-of-control hire who upsets clients, jeopardises the reputation of the organisation, and requires frequent performance management.

Harnessing the power of AI technology for good, rather than letting it run rampant through the organisation, begins with identifying the right data and having a plan for how it will be used.

It's unlikely that a single data set will yield all the information virtual agents need to understand and respond appropriately to more than the most rudimentary of enquiries.

Feeding virtual assistants with data from a disparate variety of sources will up the odds of their learning and ability to respond to queries with some of the nuance and tact that comes naturally to their flesh and blood equivalents.

When implementing AI, plan for using the right data; your IT system needs consistent monitoring and revisiting of AI strategy.

As AI in its infancy is merely a passive tool, just like the stone age tools which needed a human brain and periodic intervention to function to its fullest potential.

Easing into AI-driven solutions

AI holds enormous promise for organisations with an emphasis on user experience and customer expectations.

It starts right from automating mundane tasks increasing productivity, to reducing costs and supplying an intuitive product—provided it's introduced thoughtfully.

As the newest kid on the high-tech block, AI-powered solutions need to be monitored and managed to ensure they become part of the silent backbone of an efficient corporate enterprise, not rogue actors whose actions require regular remediation.

This initial investment in a long-term relationship with AI will ensure that an organisation's investment delivers value rather than the opposite.

 As implementation of AI in businesses increases, it will grow to be more of a generative and intuitive tool rather than a passive one.