Using AI Ethically and Responsibly
Gen-AI tools, including ChatGPT, BARD and DALL-E have revolutionised the way we generate content. They enable users to tap into the limitless possibilities of machine learning algorithms.
While the potential of GenAI is awe-inspiring, we must tread responsibly and mindfully when using these tools. Businesses should take steps to ensure their staff are educated on responsible AI usage and that there are guidelines in place.
We have produced a 1 minute video on responsible AI use (the link can be copied and shared within your organisation):
Using AI Ethically and Responsibly
Remember, there are 4 key points to bear in mind when using AI tools:
1. Only use Gen-AI in an ethical and lawful manner
Under no circumstances should AI be used for illegal or discriminatory purposes, nor should it be used to make employment decisions. If you come across AI misuse within your organisation, it is important that you report this to a manager.
2. Intellectual Property and attribution
AI-generated content cannot be protected as intellectual property. There must be some ‘independent intellectual effort’ that is directed to the form that the work takes for this to be possible. It is also imperative to recognise the possiblity that AI output may include elements of another's intellectual property. Users therefore need to avoid misrepresenting AI-generated content as their own.
3. Sharing confidential or personal information with Gen-AI
Everything that you share with an AI tool will become public. Personal, sensitive, or confidential data should therefore never be entered into AI tools. Users should always exercise caution when uploading documents and should avoid pasting any sensitive material into these tools.
4. Relying on Gen-AI Outputs
Gen-AI is not infallible. The data provided by AI may sometimes be outdated or inaccurate. Gen-AI can craft very convincing narratives, that may in fact be inaccurate or not based in reality, which are referred to as ‘hallucinations’. When assigning tasks to AI tools, you need to be explicit in your instructions, and then review the output to ensure its accuracy. AI-generated answers should not be assumed to be factual, but rather they should be used as a starting point for further refinement.