The Deloitte AI Institute has put out a report on the state of AI in the Middle East.
TL;DR – FOMO is driving investments. Talent & tech are hindering scaling up organizations’ AI pilot programs. A third of orgs investing in AI are expected to not see any ROI. Some points for vendors.
GenAI – Horses for Courses
Most orgs are using GenAI for content generation & summarization. The leading use cases per country:
UAE: Content generation
Qatar: Coding
KSA: Summarization
I find this difference to be fascinating. Is there a larger story around overall objective and culture here?
Top AI Use Cases
The functions with the most “at-scale” AI implementations are Ops, IT/Digital & Strategy. It’s surprising to see Marketing, Sales & CS at the lowest (14%). It’d seem that the use cases for which we’ve the most vendors in the market (Marketing, Sales, CS) have the least at-scale implementations in the Middle East. It’d have to do with lack of talent or strategy … or perhaps focus/investment?
Vendors: Follow the money?
The Value of AI
Value is in the eye of the report reader. Most orgs seem to be tracking employee productivity as a measure of the AI ROI with UAE focusing a bit more on non-financial benefits such as increased innovation & improved employee satisfaction.
Vendors: How well are your success stories documented?
The Three Biggest AI Challenges
The three biggest barriers to AI development & deployment include
- Choosing the right tech: understandable for an emerging tech space
- Lack of tech talent: related to the above from an implementation perspective but also perhaps from a strategy thinking perspective as well.
- Difficulty in use case identification: stems from the above two
Vendors: All the above three are to do with risk mitigation. Is your tech built on a reputable stack? Do you’ve success stories or partnerships with thought leaders through which you can help prospects identify the right AI use cases? Do you’ve the right pool of talent that can then deliver on those use cases? How much of this is demonstrably trustworthy?
Dude, where’s my data?
Data management concerns. There’s a story to tell here & I love how different stories can be told from the same data. Figure 12 on page 22 is about how managing data security-related issues & complying with regulations are of the highest concern.
It’s interesting though that that using own proprietary data is of higher concern than using sensitive client and PII data. 🙃
Mirror mirror on the wall
who’s the fairest LLM of them all?
The key criteria for model selection seemed to be whatever’s the least risky to adopt (support/access to expertise) followed by accuracy and speed.
Other AI “Demand” Related Salient Points
Investment in AI outpaces the globes despite difficulty in quantifying value. FOMO & government initiatives are driving that.
Software ate the world & now AI is going to eat your infra 🙂 It’s a good time to be in the business of providing hardware, data management, cybersecurity & communication.
Conclusion
The above is based on only half of the report. Please do go have a read to understand the state of AI in the Middle East. It would be interesting to get a better understanding of the nature of the respondents in terms of industry, organization size, market share etc.