The EU AI Act introduces significant changes to the regulation of AI in legal contexts, particularly emphasizing human oversight.

Here are the key points:

1. Human-Centric Approach: The Act mandates keeping humans in the loop, especially for high-risk AI systems. This approach aims to ensure AI systems are deployed safely and reliably, benefiting humanity while protecting human rights and dignity. 

2. Effective Oversight: AI systems must be designed to allow for human control and intervention. Article 14(1) requires high-risk AI systems to be designed for effective oversight by natural persons during their use. 

3. Risk Mitigation: The goal is to prevent or minimize risks to health, safety, and fundamental rights. Article 14(2) specifically aims to mitigate risks of high-risk AI systems infringing on fundamental rights. 

4. Competence Requirements: Human overseers need proper training and authority. Recital 48 of the AI Act proposal requires overseers to have the necessary "competence, training, and authority to carry out the role." 

5. Judicial Independence: AI cannot replace human judges. The Act emphasizes that decisions must ultimately be made by humans, particularly in legal contexts. 

However, the Act's current form leaves some questions unanswered:

- The timing of human intervention is not clearly defined.
- The standard for "meaningful" human oversight lacks clear guidelines.
- The responsibilities of human overseers are not thoroughly detailed.

As the AI landscape evolves, it's crucial for legal professionals to stay informed and engaged in shaping AI's role in justice systems. The balance between AI efficiency and human judgment in legal decision-making remains a critical area for ongoing discussion and refinement.