AI Policiies

[Sean]

Does anyone have school policiies regarding the use of AI in schools, particularily for staff who may be using AI for student reports and the risk of data being public. I assume that all the free versions won't guarantee safety of data.

[Jono Green]

This is probably a much bigger topic than the initial ask here. Ostensibly the question you seem to be asking is almost an updated Acceptable Use policy regarding IT and Cloud Applications and this is an area most schools will struggle with in terms of the split across Google and Microsoft 365. In the Microsoft 365 space, there are a number of areas that can be targeted to reduce the likelihood that sensitive data will be shared with AI applications. There are things like Defender for Endpoint and Defender for Cloud Apps which historically have provided insight into Shadow IT that will now also provide insight into Shadow AI (tm). There are also controls with Entra ID regarding how users can consent to an application to have access to data on their behalf which should be addressed as well. Microsoft's guidance for Copilot has solely been on Data Security (specifically Information Protection and DLP) which doesn't address some of the other areas that need attention, let alone the fact that nearly every school will struggle with capability and capacity to implement Information Protection and DLP.

[James Andrewartha]

Here's a collection of various schools' AI policies
https://drive.google.com/drive/folders/12CkzAGozF7uBFf3irvSYmSoA7EympJbK

Most AI services don't allow use by under 18s - ChatGPT is the notable
exception. Microsoft Copilot (Bing) doesn't support data protection for
K-12 but can be used by personal Microsoft accounts for 13+

https://support.microsoft.com/en-au/topic/copilot-in-bing-our-approach-to-responsible-ai-45b5eae8-7466-43e1-ae98-b48f8ff8fd44
https://learn.microsoft.com/en-us/copilot/faq
https://help.openai.com/en/articles/8313401-is-chatgpt-safe-for-all-ages

--
James Andrewartha
Network & Projects Engineer
Christ Church Grammar School
Claremont, Western Australia
Ph. (08) 9442 1757
Mob. 0424 160 877

[Sean]

Thanks Jono for advice and James for the resources. It is definately a huge area to be concerned about. There is a small amount of info on Netsafe, and limited info on the privacy commision website covering this area. Best advice as Jono has stated is to use the tools available to us with endpoint etc.

[David Jurgensen]

After fighting to get a policy in our school for well over a year, we finally rolled out one yesterday arvo.  Below is the relevant snip from our policy doc:

<start>

Guidelines for the responsible use of Generative Artificial Intelligence Services
Generative Artificial Intelligence (AI) tools (such as Chat GPT, Microsoft’s Bing, and Google’s Bard) are covered by the Privacy Act 2020.  Concerns raised by the office of the New Zealand privacy commissioner regarding the use of AI tools means it is important to include safeguards in our social media policy to ensure the protection of privacy and to mitigate potential risks. Before you use AI tools, please read the Ministry of Education and Privacy Commissioner's Guidance on Generative AI

In summary, AI tools pose privacy risks, particularly as they are models that learn from the data that they collect.  
1. Transparency: All content produced through any AI tool must be clearly labeled as such, and the use of AI in generating documents should be transparent.
2. Data privacy: ensure that any personal or sensitive data is handled with appropriate care and that any information processed through AI tools is done so in compliance with the Privacy Act 2020 and any other relevant legislation.  Where possible, personal and sensitive data should not be used.
3. Fairness: ensure that AI-generated content does not discriminate against any individual based on their protected characteristics, such as race, gender, age, or disability etc.
4. There is a risk that information could be exposed or misused, either through a security breach or by unintended parties gaining access. Staff are not to enter or use proprietary data or confidential information, including students information in AI tools.  While such tools are sophisticated and can provide helpful insights, they are not inherently equipped to handle sensitive data.

In summary, staff are required to exercise caution when using AI tools. Sharing proprietary and confidential information with AI poses security, confidentiality, and privacy risks.
<end>

The "Ministry of Education and Privacy Commissioner's Guidance on Generative AI" reference can be found here:
https://www.education.govt.nz/school/digital-technology/generative-ai/

And below is a useful resource to have handy too:

https://www.hcamag.com/asia/specialisation/hr-technology/will-ai-use-put-your-organisation-at-risk/466635

[Andrew Hood]

Hi,

I was one small part of the large team that put together the Ministry advice page on AI that David linked to. (Bumping the link here:  https://www.education.govt.nz/school/digital-technology/generative-ai/)

If there is feedback or areas of concern that you think we should cover in the Ministry advice, please feel free to Email digital...@education.govt.nz. We cannot offer you policy reviews, give you App by App detailed configuration advice or offer legal advice on every T&C's out there but we may be able to provide some general guidance, point you in the right direction or help us update and modify the advice on the web page.

In general, my advice would be:

• Don't put student data into any AI platform unless you fully understand the risks
• Don't use AI where the outcome could have significant outcomes on others unless there is an adult review step in the process, an even then, progress with caution
• The public perception to AI is rapidly changing - so when using AI, consider the range of public reactions that may be invoked from doing so.
• Not all AI is bad and AI can offer significant process and efficiency gains
• I have seen some great demonstrations of AI being used in a pedagogical space to improve student outcomes, but understand the controls in place before replicating those things
• Talk to others who may be doing similar work to get their advice and feedback - I saw a great talk from Tony Gilbert at NewEra the other day on AI in Schools

Thanks,

Andrew