Part 3: Ethical Considerations in Artificial Intelligence, not just in Safety. Safely mixing real intelligence with AI - its coming to a workplace near you – where is the ethical line?

Bias and Fairness: AI systems can accidently, or intentionally, cause a bias, leading to unfair outcomes. For example – what if the only car you had ever seen was a blue four door Honda.  Every driving decision you make would be around that type of car because you had never “seen” anything else. You would of course teach a computer the same bias.

Privacy Concerns: Theses are significant because of the collection of personal data for AI training.  What if one person followed you everywhere you went and recording everything.  How good would that person be at predicting your next move or what you would like to eat or buy?  What if it wasn’t one person but 100 people.  Could they get better at predicting?  What if it were infinite numbers of people (AI computers) with all of your history?  All that data being collected about someone that is being stored forever for some purpose in the future that we can’t even imagine today.

What if there was some health data in your digital world that made you miss out on a job?  What about details from a jilted ex – and that made you miss out on meeting the spouse of your dreams?  Maybe a mental health issue that you overcame?  What about a younger reckless you versus the current mature you – judged by your peers much???  So … Should an infinite, and possibly intimate, database be alarming?  Part of the concern could lie in the fact that it always stays recorded and could always be accessed from anywhere by anyone and used for any purpose, forever.

Regulation and Governance: As AI technology evolves, there’s an ongoing debate about creating effective regulations to ensure ethical usage only. However, you must realize by now that regulation has yet to catch up with emerging technologies and that it may not ever.  This means that company leaders will need to set up protections and rules around the use of AI, to ensure protection of their company’s reputation, before there are government rules in place.  Again, once the data exists it exists forever, and that history could come back to haunt your company. Think “Ford Pinto” (Ford thought a small part was too expensive to replace and computed the cost amount of death lawsuits against cost of the replacement part.  Turns out that calculation was wrong - both in court and in hindsight).

Conclusion:  AI already exists in your everyday life.  Data about your workplace safety is already being recorded and stored (think things like driving habits). AI is around you and you may not have even know it, and it is growing fast.  There are a lot of examples of how companies are using computers to simulate human interactions and guide desired outcomes. Negatively - It can feel like you are being manipulated, or Positively - it can feel like you used the technology to come to the best decision.  Either way – a lot of data is being stored!  

Workplace safety can mean everything from physical to psychological and ethical considerations will always be an overarching umbrella for your organization. No matter how much you learn about AI, from a practical business point of view, it is important when you go to use it to start small and learn to walk before you try to run (so you can recover from an unplanned failure).  

The Plug - Workhub’s simple, all-inclusive $4 per user per month Compliance management software has a number of functions (like surveys, suggestions and polls) to help you engage with your employees to so that you can work together to work forward on understanding safety and the impact it has on everyone. Our customer support team is always available to help and its always free to book a demo at www.Workhub.com.

This is the third article of a five-part series on Safely mixing real intelligence with artificial intelligence. You can find the rest of the series at our website under “blog”.

Part 1: Introduction to AI in Safety

Part 2: How AI Works in Safety

Part 3: Ethical Considerations in AI, not just in Safety

Part 4: AI in the Workplace Safety Program

Part 5: The Future of AI in Safety