A compliance audit of your AI tools
A legal review of the AI tools you use or develop.
Govern your use of artificial intelligence, protect your data, and respond quickly to a leak or a cyberattack.
In the event of a data leak or a cyberattack, contact the firm without delay so that the first legal decisions can be framed.
Here are the most common situations.
Your colleagues use AI models or agents and may enter customer or employee data, or confidential information, into them.
Personal data of customers, employees or partners has been lost, disclosed or accessed without authorisation.
Ransomware, an email account takeover or an intrusion affects your systems, and you must manage the incident and its legal consequences.
You design an artificial intelligence tool or build it into your product, and it processes personal data.
An AI provider processes your data on servers located abroad.
You entrust your systems or data to a hosting provider, a software vendor or a processor, and you want to govern their security commitments by contract.
Automated processing evaluates people or helps make decisions, for example to recruit, score or offer a service.
You provide services to European customers, or the outputs of your tools are used in Europe.
You want a clear procedure in place before an incident occurs.
You are a public body, or you act on its behalf, and specific rules on information systems security may apply to you.
Here is what they provide, explained simply.
There is currently no Algerian law specific to artificial intelligence. AI tools fall under general law, in particular Law 18-07 as amended and supplemented by Law 25-11, and Law 26-02 on trust services.
The controller must implement technical and organisational measures suited to the risks in order to protect personal data.
A breach is a security incident that leads to the loss, destruction, alteration, disclosure of, or unauthorised access to, personal data. You must manage the incident, document it and meet the notification obligations that apply.
The controller may transfer personal data to a foreign state only with ANPDP authorisation, and if that state ensures an adequate level of protection. Derogations exist, for example with the person’s express consent. When an AI tool processes data outside Algeria, the situation must be analysed case by case.
The law defines profiling as any use of personal data to evaluate certain aspects of a person, such as their performance at work, economic situation, health, preferences or behaviour. Since Law 25-11, the rule governing decisions taken solely on the basis of automated processing appears among the provisions specific to processing for criminal law purposes. Its scope for other processing must be assessed case by case.
The controller must choose a processor that offers sufficient guarantees, and govern its work by a contract or other legal act.
The controller appoints a data protection officer and keeps a record of its processing activities. The controller and its processor each keep an automated log of operations, which makes it possible to identify the reasons for, and the date and time of, the operations.
The ANPDP receives complaints, may inspect the places where processing takes place and access the data, and may impose administrative sanctions.
An Algerian company may be affected by the European AI regulation and by the GDPR, in particular where the outputs of its tools are used in the Union or where it offers services to people located in the Union. The heaviest obligations of the AI regulation concern high-risk systems.
Unregulated use of artificial intelligence or a poorly managed incident exposes your organisation to several kinds of consequences.
The law provides administrative sanctions, imposed by the ANPDP, and criminal penalties, in particular for transferring data abroad without authorisation.
Data entered into a tool that is not properly controlled may become known to third parties, which may engage your organisation’s liability.
The people affected may lodge a complaint with the ANPDP, and the incident may also lead to disputes with your customers or partners.
Opaque use of AI or a data leak weakens the trust of your customers, employees and partners.
Our role is legal and organisational. We support you in four steps, before as well as during an incident.
We identify the AI tools your teams use or that you develop, and the data they process. We review your processing, your providers and your current procedures to measure your legal exposure.
We examine the rules that apply to your uses, data transfers and, if you target the European market, the AI regulation and the GDPR.
We draft your usage policy and adapt your contracts with AI tool providers. With you, we draft the incident management procedure, the allocation of roles and the templates for decisions and notifications. We adapt your contracts with hosting providers, software vendors and processors to include their security commitments.
We train your teams in good practice, then support you as your tools evolve. We help you classify the incident, take the legal decisions, meet the notification obligations that apply and build your file. Our work does not replace that of a technical provider.
Here is what we give you.
A legal review of the AI tools you use or develop.
Clear rules for your teams on the use of models, agents and generative AI tools.
A written procedure that sets out who does what, in what order and with which decisions to take. A template for recording each incident, its effects and the measures taken.
Suitable clauses in your contracts with AI tool providers. The security commitments to write into your contracts with hosting providers, software vendors and processors.
Templates for notifying the ANPDP and the individuals concerned, to be used when the obligations apply.
Your decisions, procedures and evidence gathered so that they can be presented if the authority asks.
There is currently no law specific to artificial intelligence. AI tools are subject to general law, in particular Law 18-07 as amended and supplemented by Law 25-11, and Law 26-02 on trust services.
The texts do not address these tools in particular. However, entering personal or confidential data into an external tool raises questions, notably on data transfers. A usage policy for AI models and agents sets the rules to follow.
If the tool processes personal data outside Algeria, the situation must be analysed under the rules on data transfers. We review your case.
You need to act quickly, classify the incident, then meet the notification obligations that apply. We help you with the legal decisions and with documenting the incident.
In some cases, the law may require notification to the ANPDP and to the individuals concerned. We analyse your situation to determine the obligations that apply to you.
No. Our role is legal and organisational. We help you manage the legal consequences of the incident and liaise with your technical providers.
A first conversation to understand your situation and propose suitable support.
At Amad Advisory, data protection is our profession, and it starts here. Do you agree that we measure your visit, without identifying you? Take the time to understand before you choose.
It is any information that can identify you, directly or indirectly, for example your name, your email address, your telephone number or your IP address.
It is your agreement, given knowingly, for your data to be used. You can give it or refuse it with no effect on your use of the site, and you can withdraw it at any time.
The controller is Amad Advisory, EURL, No. 01, Résidence du Petit Port, Aïn Benian, Algiers. You can reach our data protection officer at dpo@amadadvisory.com. Only the firm reads your messages and the visit statistics. Our host in Algeria provides the technical storage, and its email service delivers your messages by email, without using them for its own account.
Without your agreement, the site measures nothing. With your agreement, it collects visit statistics, such as the pages viewed and the length of the visit, which cannot identify you. The contact and quote forms collect only what you enter in them, and only if you fill them in.
If you accept, your visit is counted in statistics that cannot identify you, to help us see which pages are useful and to improve the site. Your choice is stored in a small technical cookie, the only one the site sets. If you refuse, nothing is measured, and the site works in exactly the same way. Your refusal is stored in the same way, so that you are not asked again on every page.
No. The audience measurement is hosted in Algeria, and its data does not leave the country. If you use a WhatsApp link, you leave the site and that service applies its own rules.
Your choice is stored for six months, then you are asked again. Visit statistics are kept for one year. Messages sent through the forms are kept for one year, then deleted.
Yes. Using the « Privacy » link at the bottom of every page, you can withdraw your consent at any time. You have a right of access, rectification and objection. To exercise it, write to dpo@amadadvisory.com. You can also refer the matter to the National Authority for the Protection of Personal Data. To find out more, read our privacy policy.