Template: Artificial Intelligence and Governance Framework

More on Artificial Intelligence from ICDA

An AI governance framework can guide the board’s response to the opportunities and risks that come with AI and thus keep the organisation from falling behind or falling into danger.

Investing time and effort at an early stage to create a governance framework can help the board down the track by preventing future trip-ups and avoidable crises.

Such a framework is designed to guide the organisation in situations that involve AI (or might involve AI in the future). It’s not a policy but a set of understandings and priorities that have been agreed by the board. The board should refer to the framework when seeking to make decisions involving AI.

Download our AI governance framework template by clicking the buttons below. The Microsoft Word version will allow you to edit the template to suit the specific needs of your organisation.

What's inside?

1. Purpose of the artificial intelligence governance framework

Artificial intelligence (AI) has enormous potential to benefit community organisations. However, its use also brings risks. The board and executive of [name of organisation] is responsible for being aware of these risks and managing them in relation to decision-making processes and the organisation’s use of AI generally.

2. Scope

The AI governance framework offers guidance to the board and executive regarding decision
making about the use of AI in [name of organisation]. To mitigate risks, [name of organisation]
has established clear guidelines for the use of AI systems, ensuring human oversight, factchecking procedures, transparency in decision-making processes, and ethical considerations.

3. Key principle

[Name of organisation]’s board and staff should be aware of the limitations and potential biases of AI-generated content and should draw on human expertise and ethics as well as AI when making AI-assisted decisions.

4. Policies

  • Policies related to the use of artificial intelligence at [name of organisation]:
  • Policies related to the use of artificial intelligence commonly used in the [name sector]:
  • Current gaps in policy or procedures related to the use of artificial intelligence at [name of organisation]:

5. Oversight

The committee overseeing the use of artificial intelligence at [name of organisation] is:
[name of committee]

Members of that committee are:
[names]

The committee is responsible for identifying risks associated with the use of artificial intelligence in [name of organisation] and raising these risks with the CEO and the board.

6. The use of AI to assist board decision making

Ensuring that AI systems are trained on diverse and representative data, and that staff and board members are asking these systems the right questions, is essential to prevent bias and discrimination at [name of organisation].

Procedures in place to ensure that AI is used to enhance human judgment rather than replace it are: [list procedures]

7. Accountability in decision-making

[Name of organisation]’s board should uphold high standards of ethical behaviour in executing its fiduciary duty. It must avoid conflicts of interest and prioritise the best interests of the community over personal or special interests.

[Name of organisation] has established clear lines of responsibility to ensure the board understands the legal implications of AI deployment.

The CEO is responsible for presenting the due diligence data related to any new use of AI in the organisation to the oversight committee prior to use.

8. Board oversight

AI applications in use at [name of organisation] are outlined in the following table.

Name of AI application Purpose Users Is the app human-assisted (e.g. ChatGPT) or human oversighted (e.g. chatbots)? Do we pay for this app or is it publicly available and free? Where is the data held? Risks of this application in the way it is currently used
e.g. ChatGPT
e.g. Chatbot
e.g. calendar scheduling/email sorting
e.g. e-commerce personalisation
e.g. voice recognition and/or transcript
e.g. social media management, scheduling posts

AI applications currently used by other organisations in our sector are outlined in the following table.

Name of AI application Purpose Users Is the app human-assisted (e.g. ChatGPT) or human oversighted (e.g. chatbots)? Do we pay for this app or is it publicly available and free? Where is the data held? Risks of this application in the way it is currently used
e.g. ChatGPT
e.g. Chatbot
e.g. calendar scheduling/email sorting
e.g. e-commerce personalisation
e.g. voice recognition and/or transcript
e.g. social media management, scheduling posts

AI apps are used to inform decision making at [name of organisation] in the following ways:

[List]

9. Long-term monitoring

[Name of organisation] continuously monitors and assesses its AI systems to ensure they remain aligned with the board’s and the organisation’s goals, via the following mechanisms:

  • E.g. “AI current usage, opportunities and risk assessment” is on the board agenda as a standard item twice a year.
  • E.g. the board is confident that staff are aware of ethical considerations related to AI use, such as data privacy and bias implications, because the CEO reports on this regularly.

10. Sustainability and risk management

[Name of organisation]’s board exercises due diligence in overseeing the organisation’s operations, finances and risk management relating to AI usage. The board is confident that use of AI as a mode of saving time and money has been adequately explored by the professional staff team. It knows this because:

  • [X]% of staff require training on AI technology to achieve a basic level of competency.
  • [X]% of staff require training on AI technology to achieve an ideal level of competency.

The latest AI training plan for staff is outlined below. [Presented to the board in summary form if requested; the CEO has oversight.]

Staff role AI apps required for ideal productivity Training required? Training completed

More on Artificial Intelligence from ICDA

Become a member of ICDA – it's free!