In many Algerian companies, generative AI has already crept in through the back door: one employee drafts a letter with ChatGPT, another summarizes a report, a third translates a memo. The question is no longer whether to open the door, but who sets the framework and how teams learn to use it properly.
On this point, executives and managers from the diaspora have something useful to share: recent first-hand experience of the mistakes to avoid, gained in European companies that had to establish rules quickly. The technology itself is the same in Algiers as it is in Paris.
Start from tasks, not tools
The training programmes that work least well present a tool, its menus and its promises. A week later, participants have no idea what to do with it on Monday morning. It is better to start from the list of real tasks within a given role. For a claims manager, those are letters to policyholders and case summaries. For a bank relationship manager, responses to complaints and meeting preparation. For an HR team, job descriptions and interview notes.
The Afriques newsletter
- Free & no spam
- 1 email per day
- Unsubscribe in 1 click
You take those tasks one by one. You look at what AI can prepare, what the employee must check, and what should never be entrusted to it. The benefit is measured on a specific document, not on a general impression. Each profession leaves with its own examples, in its own working language.
A data rule, written and adapted to Algerian law
In Europe, companies relied on the GDPR. In Algeria, the reference text is Law No. 18-07 of 10 June 2018 on the protection of natural persons in the processing of personal data, as amended and supplemented by Law No. 25-11 of 24 July 2025. This reform notably requires the data controller to appoint a data protection officer and to maintain a register of processing activities. The law also subjects the transfer of personal data to a foreign state to authorisation by the ANPDP.
Yet the most widely used generative AI assistants are online services operated by foreign companies. Pasting a client’s name, account number or the contents of a claims file into one of them is far from trivial. The question of data transfer deserves to be raised with the officer before the habit becomes established.
The practical answer fits on one page: which tools are authorised, which data must never be entered into them, who reviews a text produced by AI before it is sent, and to whom a concern should be reported. A short rule known to everyone offers better protection than a prohibition that no one respects. It is validated internally, with the data protection officer and a lawyer familiar with the text.
Designated contacts within each department
Training everyone at once is costly and quickly forgotten. A more sustainable approach is to designate one or two contacts per department. These are curious employees, not necessarily IT specialists. They receive more in-depth training, keep the examples for their field up to date, and become the first point of contact for their colleagues. Questions flow up through them, and best practices flow back down through them.
This arrangement has another advantage: it builds skills that remain within the company. A contact trained in Oran or Constantine does not need an external trainer to help their team progress the following year.
The role of the diaspora
Many Algerian executives based abroad have experienced these early stages from the inside: the first improvised uses, the hastily written charters, the failed training sessions that were later corrected. They can share this experience without trying to impose it. The legal framework is Algerian, as are the professions and the priorities. The method, however, travels well: start from tasks, write the data rule, rely on internal contacts.
Algeria is investing in training new talent in artificial intelligence. There remains an audience that receives less attention: employees already in post, who are using these tools today. Training them in safe usage may seem a modest undertaking. Yet it is on this that a large part of the answer depends: will generative AI save companies time, or will it expose them to risk?
Aïssa Khelladi, Franco-Algerian, founder of Skillevos, a generative AI training organisation Qualiopi-certified in France.
Aïssa Khelladi
Aïssa Khelladi is the founder of Skillevos, a Qualiopi-certified generative AI training organisation and France Num activator. Franco-Algerian, he helps companies upskill their teams in AI job by job, and stay compliant with the AI Act and GDPR. A columnist for the Journal du Net, he previously spent fifteen years in B2B sales to HR departments and key accounts.
View all articles → · LinkedIn ↗ · skillevos.fr ↗
