Today, the Ministry of Transport, Communications, and Information Technology issued a new policy on government use of AI technology. Similar to other policies issued by the MTCIT, this document is not legally binding it is essentially a form of instructions and recommendations for government entities.
The policy appears to have been finalised last June and is addressed to government entities only. It sets out six very high level principles for the use of AI: inclusiveness, human-centrism, accountability, fairness, transparency, and safety.
The key concrete policy provisions include:
- A requirement to undertake a risk assessment in regard to the use of AI (but no details as to when such a risk assessment should be done, if it is a one-off requirement or a periodic requirement).
- A requirement to ensure that no AI bias takes place and disclose to those affected by the AI system of the existence of such a system and the impact it may have on them.
- A requirement to obtain prior approval of those affected by the AI system.
The policy also briefly mentions a requirement to comply with open data policies.
You can read the Arabic version of the policy on this link.