Page 393 - CyberSecurityEssentialsEnglish

Communities of practice help spread expertise without excessive centralization.
Representatives from the units meet to share failure cases, templates, evaluations, and
vendors. These communities should focus on transferable evidence and experience, not only

on success stories; knowing what did not work may save an organization more time and
money than a success that cannot be repeated.

Sustainability and Resources
Resource impact belongs in the adoption decision, especially where large models are used

repeatedly or local infrastructure is run. A reader does not need to compute exact
consumption for every request, but should ask whether the model in use is larger than the
need, whether caching, batching, or a smaller model would serve, and whether the value

justifies the volume of use. Technical, financial, and environmental efficiency usually move
in the same direction.

Symbolic sustainability targets that nobody measures should also be avoided. A fair

comparison looks at the whole process: a system may reduce travel, rework, or time, while
adding computation and storage. The better decision rests on vendor or internal
infrastructure data and on the actual volume of use, and it is reviewed when the system

scales.

The Arabic Context: Global Principles Need Local Translation
Arab organizations face the same global questions, but the context adds important
considerations: the quality of Arabic and its dialects, bilingual documents, where data is

hosted, national and sectoral laws, uneven digital skills, cultural sensitivity, and reliance on
global vendors who may not understand local procedures.

Several countries in the region have published national resources. Saudi Arabia has issued AI
ethics principles and frameworks and guidance for government adoption and use, which
helps turn principles into assessments and controls inside the national context [10]. The

United Arab Emirates gathers official resources including an AI ethics guide, guidance on
synthetic media, and a national strategy [11]. Egypt has announced the second edition of its

national AI strategy for 2025 to 2030, with a vision for a comprehensive framework
supporting digital development and economic and social growth [12].

The existence of a national strategy does not mean an organization has obtained a ready-
made policy. It must identify the laws and regulators that apply to it and connect them to its

sector, its contracts, and its data. It must also test performance in Arabic on real tasks, rather
than settling for an Arabic interface or attractive answers.

392