Forty-five percent of large Moroccan enterprises have adopted generative AI tools, according to LNT.ma. That is fifteen percentage points higher than six months ago. Meanwhile, SNRTnews reports a brutal shortage of AI experts. Recruitment firms like Jobzyn and Maki cannot keep up. Al Akhawayn University is convening CEOs and HR leaders to address the gap.
We are repeating the ERP mistakes of 2015. Back then, we bought software without training users. Today, companies deploy ChatGPT Enterprise and Microsoft Copilot without knowing who will configure, secure, and integrate them into business workflows. The result is predictable. Expensive licenses used to draft emails slightly faster. That is the extent of it.
The AI recruitment market in Morocco has become a jungle. A data scientist with three years of experience gets salary offers doubled by competitors. Banks like Attijariwafa and BMCE, plus telecoms like IAM and Inwi, are poaching talent from startups and IT services firms. At Xpertize, we see impossible briefs. “I need a prompt engineer, bilingual Arabic-French, with DevOps culture and banking sector knowledge.” This profile does not exist. Or rather, it exists in Casablanca for forty thousand dirhams monthly. Minimum.
Schools cannot keep pace. INSEA and EMI graduate two hundred data scientists per year. The market demands two thousand. The gap between classroom and office floor is massive. When a student knows Python machine learning, they often ignore how to secure customer data flows in a Moroccan bank subject to AMMC regulation. Al Akhawayn University understands the issue. But they do not train fast enough.
In our service centers in Casablanca and Rabat, we see twenty-five percent productivity gains on candidate processing when generative AI is used correctly. But this requires trained operators. Not just on the tool, but on verifying hallucinations. An AI recruiter who cannot spot a CV inflated by ChatGPT is dangerous.
Our BPO business is hit hard. European clients demand AI in our processes. They want automated invoice processing, lead qualification. We must recruit prompt engineers for salaries of twenty-five to thirty-five thousand dirhams. That is three times a senior operator’s salary. This changes the economic model of Moroccan nearshoring.
If I were you, I would stop hunting for the perfect profile. It is not on the market. Look for hybrid profiles. A good salesperson who learns prompt engineering beats an AI engineer who understands nothing about business.
Stop buying tools before training teams. Use the 70-20-10 rule. Seventy percent internal training, twenty percent external mentorship, ten percent spot recruitment. Moroccan enterprises do the opposite. They buy the tech, then call to find a savior. He will not arrive.
Sign partnerships with schools now. Not in two years. Jobzyn had the right idea with Al Akhawayn. Generalize this model to EMI, INSEA, HEC Morocco. Long internships, real missions, concrete cases. Not academic projects.
Look East for sourcing. Morocco cannot produce enough local AI talent before 2028. Import, train, and retain. The Moroccan diaspora in Montreal, Paris, and Berlin is a lead. They know the local context and international standards.
Generative AI is not a purchasing decision. It is a talent decision. Forty-five percent adoption means forty-five percent frustration if we do not change methods. The productivity gains McKinsey promises will not come from Microsoft licenses. They will come from people you cannot recruit today. Start there.