1. AI companies:
These are the people creating AI tools and profiting from them. They can be seen as responsible for any problems that arise from AI use. Changes made to AI will have to come from them, and thus, solutions must compromise with them. As the source of the issue, they may have the largest stake in terms of their power to decide what the solution should be.
2. Workers who use AI:
These are the everyday people using and interacting with AI models such as chatbots. Though they may not be the most at risk, changes made to AI tools may affect their work. If an AI model is changed and given the ability to push back on what it is told, this could result in difficulty using it as a tool for work if changes are implemented incorrectly.
3. People with mental illnesses:
People affected by mental illnesses like schizophrenia, OCD, or others are at risk of delusions or intrusive thoughts. AI can cause these people harm by feeding into these thoughts and encouraging them. If a person presents a delusional idea to AI, it is not currently equipped to claim that the user is wrong and help them in a medically accurate way. These people are most at risk from this aspect of AI, and solutions must be found to negate this risk.
4. People who care for those with mental illnesses:
The family and friends of those who are at risk of AI psychosis. These people can be affected in a number of ways by AI and its ability to cause psychological harm. Much of the initial care given to people with mental illness will likely come from friends or family and the responsibility of care can cause stress.
5. Environmentalists/the environment:
Any AI-related issue will have environmentalists as a stakeholder group. AI’s environmental impact cannot be understated. Any improvements to make AI more widespread and effective will also come with more expansion of the industry and thus more environmental destruction.