1.   Where automated models are used for assessing credit risk of crowdfunding projects or project owners and for approving crowdfunding projects to be proposed to investors, crowdfunding service providers shall ensure that:

(a)	the roles and functions concerned have a good understanding of the methodology of those models, their input data, and their assumptions and limitations;

(b)	the management body has a sufficient understanding of the use of technology-enabled innovation applied to financial products;

(c)	the automated models are fit for purpose, and that their use is proportionate to the size and complexity of the activity of the project owner, of the crowdfunding project and the amount of the loan.

2.   For the purposes of paragraph 1, crowdfunding service providers shall have in place policies and procedures and set out appropriate governance arrangements for the design and use of such automated models.

3.   The policies and procedures referred to in paragraph 2 shall:

(a)	ensure the quality of data used as input for the automated models;

(b)	ensure that the quality of output of the automated models is regularly assessed;

(c)	establish criteria for deciding when the outcome of such automated models can be overruled.

4.   Crowdfunding service providers shall have in place adequate documentation covering the methodology, data input and the criteria that automated models use for assessing credit risk, monitoring credit risk and approving crowdfunding projects to be proposed to investors.