Using AI in aXcelerate
A guide to best practices and principles for using artificial intelligence (AI) effectively in aXcelerate
Our mission for introducing AI into aXcelerate is to empower training organisation teams to drive operational excellence.
At aXcelerate, we believe AI should be used to empower educators and training organisation staff to have more time to focus on what they’re passionate about – delivering high-quality training that sees learners succeeding in the world of work.
Our team has always been led by the HEART values – honesty, empathy, acceptance, respect and trust. We will continue to use these values to guide our approach to AI.
As we move forward, we’re committed to working in partnership with our clients and partners to achieve this vision for implementing AI in aXcelerate.
Our AI Principles
AI should be used to streamline processes, empowering individuals and teams to work smarter, boost productivity with confidence and achieve their goals efficiently.
Generative AI tools should be used in ways that are open to challenge and retain human agency and accountability for decisions. People always have choice, control and accountability over the role of AI in their work. AI should always be used in a safe, accessible, fair and respectful way.
We prioritise safety and security in our AI approach, ensuring that all stakeholders utilising generative AI tools in aXcelerate benefit from robust privacy and data protection measures.
We commit to maintaining transparency regarding our partnerships, security, and data collection, protection and management practices. Our AI partners do not use your data to train their models.
AI considerations and best practices
AI can be a powerful tool, and it’s important to know how to use it well.
Accuracy
AI-generated outputs and results are not perfect, and must be checked for accuracy. AI uses artificial intelligence algorithms and data collected from the internet. There is a possibility of inaccuracies, omissions, algorithmic bias or misleading outcomes due to the complexities and limitations of AI technologies. All outputs generated by AI in aXcelerate should be checked by humans.
Using generative AI in Online Assessments
aXcelerate’s generative AI capabilities have limitations. It’s important to ensure you are not solely relying on AI-generated question outputs without checking and considering ASQA’s Principles of Assessment. Human judgement, empathy and nuanced decision-making remain essential in the assessment and lesson authoring process.
Validity
You must ensure that assessment tasks and methods match assessment requirements.
ASQA example: If assessing a practical skill such as keyboarding, questions about how a keyboard operates may not be valid as this knowledge is not required in order to carry out the task. Instead, use questions that demonstrate knowledge of why the student is doing the task in a particular way.
AI consideration: If an assessor inputs a PDF on keyboarding, the AI generator may output a question about how a keyboard operates. The assessor needs to take this into account and review the instructions they are inputting, as well as review and adjust the output.
Reliability
Training organisations should have a well-designed assessment system that includes measures to minimise variation between assessors. Training organisations should also develop evidence criteria to judge the quality of performance in order to help assessors make consistent judgements about competence. Where evidence criteria includes model answers, assessors need to be involved in choosing to use or overwrite AI-generated suggested responses.
Your responsibilities
- By using AI features in aXcelerate, you agree to also follow OpenAI’s usage policies.
- You should exercise your own judgement and discretion when using AI-powered tools. It’s essential to verify and cross-check the information provided by the tool before making any decisions, taking actions, or relying on it for critical matters.
- You should let others know when you’re using or presenting AI-generated outputs. aXcelerate does not currently indicate when an output is AI-generated within the platform.
View aXcelerate's Beta Program: Additional Terms and find out more about our Beta Programs.
Data usage and privacy
aXcelerate’s agreement with OpenAI provides that data is not to be used for model training. Data inputted into AI features is sent to OpenAI for processing and generating AI output. Data is temporarily retained in OpenAI’s environment to provide the services, however it is not used for model training. Here’s an example of how the features work at a high level:
- You input or upload text in the Generate Questions form as a prompt. You can also add some context about the action you want to achieve, such as generating a Multiple Choice question from the text.
- The prompt text is sent to the OpenAI API along with some context about the action you want to achieve, like asking for a summary of the selected text or more information on a subject.
- OpenAI returns a text output that attempts to match whatever instructions or context are included with the original input.
- The output is written back to aXcelerate as an assessment question.
Visit the OpenAI Security Portal to learn more about data retention and security practices.
Visit aXcelerate’s Terms of Use to learn more about your data in aXcelerate as a current client.
