Even though there have been significant investments by Middle East governments in artificial intelligence initiatives, and some pioneering enterprises have rolled out AI applications, growing c-suite interest has not yet resulted in the technology being broadly deployed.
A handful of countries have taken the lead in AI-related spending, adoption and education initiatives. Much of the spending on AI technology is happening in key markets such as the UAE, Saudi Arabia and Turkey, according to IDC.
In the UAE alone AI is expected to contribute US$96 billion to the economy by 2030, according to PwC. On its part, the Egyptian government has funded several initiatives to spur AI deployment and aims to have 7.7 percent of its GDP derived through AI by 2030.
Yet broadly throughout the Middle East, actual AI deployment is lagging behind general interest in the technology.
Even though overall spending on IT in the region is expected to increase 1.8 percent to US$160 billion this year, “only a few leading local organizations are overcoming technology hurdles, and moving more quickly toward artificial intelligence and digital business systems,” said Gartner analyst John Lovelock in the company’s 2019 regional forecast.
Actual implementation is lagging even in those countries that have spent sgnificant funds fostering AI innovation. For example, only 25% of UAE businesses have a fully implemented AI strategy, according to an Avaya-sponsored report by Vanson Bourne.
The region’s CIOs are keen to adopt more solutions that make use of AI and a vanguard of enterprises have rolled out AI-based applications, saying that if they don’t get on board with AI now it will cost their organisation for the next decade. So what’s holding other enterprises back?
Lack of AI understanding around AI
There are several barriers to adoption, some specific to the region and others challenging CIOs all over the world.
A lack of understanding of the technology itself is a big stumbling block. The reason behind this is at least partly down to the fact that the market is young, said Arup Roy, research vice president at Gartner. “You have to appreciate the fact it’s a new area. There are hardly any references, useable case studies, and this is a big impediment,” Roy said.
Many organisations don’t know where to start, often implementing AI for AI’s sake without taking the time to research whether there’s actually a business case for the technology, Roy noted.
“The first step for the end user organisation is to decide where and why you should be implementing AI,” said Roy.
“Use cases, experience and precedent are missing. The CIO isn’t actually very well equipped to do this business preparation so you need an expert to do the business case and ROI development,” Roy said. “Without this, if you just follow the hype you could get someone in to do a proof of concept (POC) and discover you won’t get the ROI. The project then never moves into production.”
Another scenario is when organisations move a project into production and realise that this technology is an “entirely different ball game”, Roy said.
There may be a variety of reasons why something that works during the proof of concept stage may not work when it’s deployed to end users. When a project runs into problems in a production environment, the confidence level in the technology drops and the business may drop the initiative entirely. “It’s early days, as we move on the journey a lot of learning will be captured and feed into projects. But all this takes time, say 3-5 years,” Roy said.
AI skills gap constrains rollouts
Resource constraints are another issue holding businesses back – 42% of those questioned by Vanson Bourne said they lack the in-house skills to enable adoption.
Many of those with the skill sets businesses need to implement AI are fresh out of university – new graduates with minimal, if any, real-world project experience. Then, if you’re actually able to find those resources they’re expensive to recruit and hard to retain, Roy noted
Demand for those with AI knowledge is high, therefore the power is in the hands of the professional. This allows them to press for high salaries and take the next best offer when it comes along. The natural inclination for businesses in the region has therefore been to rely on external providers rather than recruit for in-house skills. Then it’s back to the issue of a lack of knowledge. How do organisations know which service providers are the right partners?
AI trends in the Middle East
Even with these barriers to adoption, there are of course many AI-related projects underway in the Gulf Cooperation Council (GCC). The region’s most popular applications have focused around customer experience, and in particular, chatbots. This may be due to the fact that consumers in the region are asking for them.
“Our SuperServe research found that consumers in Saudi Arabia (80%) and the UAE (82%) express a strong preference for AI-powered chatbots, which is significantly higher than the global average of 60%,” said Ahmed Helmy, CTO of Avaya International.
Adoption of other technologies may act as a gateway for businesses to move towards more complex technologies such as AI.
There is a trend toward adoption of adopt robotic process automation (RPA) and that could lead to businesses ushering in technologies with high levels of complexity, as machine learning is increasingly being embedded into RPA software, Roy said.
“Organisations are going to dip their toes in by signing up to these lesser complex technologies and this will lead them to see the benefits of things like AI,” Roy said. “This is something unique to the region.”
In the next five years, Roy adds, we will start to see more sophisticated AI applications deployed in the Middle East, much of it driven by government bodies and agencies, as initiatives come to fruition and an increasing number of enterprise projects highlight real-life use cases.