Why AI ethics matter: F4S applies responsible AI principles

It's hard to ignore the endless potential of AI. It has the power to increase efficiencies across all industries, by streamlining services, automating repetitive tasks, and reducing human error. However, as with any truly disruptive technology (such as the world wide web when it was first created), there's an equal amount of concern, as well as moral decisions, surrounding AI.

In the right hands, AI can cause a ripple effect of positive social change. However, the responsible use of this fast-growing technology hinges on one important point: ethics. In this article, we delve into some of the important ethical risks and questions around AI use — and what this looks like in the coaching industry, in particular. We also cover the steps Fingerprint for Success (F4S) takes to ensure we are using AI for the greatest good.

AI ethics: why is it important?

IBM defines AI ethics as a set of guidelines that advise on the design and outcomes of artificial intelligence¹. Unlike moral principles (which are typically an individual's subjective idea of right or wrong), ethical guidelines provide a code of conduct that a community can adhere to. This serves as a North Star that organizations can use to guide their decision-making when working with AI. This is especially important in the absence of AI's legal status.

Having these ethical guidelines is crucial because people are prone to cognitive biases that cloud our decision-making. This flawed human judgment can be reflected in the data we generate. And this data is then used as the foundation for machine learning. It's important that we create AI algorithms that use the best possible data, otherwise this flawed thinking may become amplified at scale.

There are other ethical issues around AI. It may replace people's jobs, lead to invasions of privacy, and even cause accidents, for example, when it comes to self-driving cars. As you can tell, the ethics of technology is important for the safety and well-being of our society.

AI ethics in the coaching industry: what is it?

So, what do AI ethics look like in the coaching industry, specifically? Because coaches work so closely with their clients, this creates a range of unique challenges. One such consideration is privacy and data protection. Because coaching often involves collecting and analyzing personal data, AI coaching companies must ensure that information is handled securely and confidentially. Maintaining proper consent from clients is an essential ethical obligation, too.

As mentioned, bias and fairness are also important ethical principles. AI systems can inadvertently inherit biases from the data they are trained on, which may lead to discriminatory outcomes. Therefore, it's crucial that organizations take a human-centered approach to training AI coaches, to ensure individuals of all demographics are treated equitably. This is particularly important in tools such as Fingerprint for Success, where AI technology transcends simple coaching but integrates with hiring and performance-management processes. As such, we take our responsibility to avoid potential risks incredibly seriously.

What happens when AI ethics aren't in place?

There are already plenty of cautionary tales about the importance of having ethical guidelines in place in the HR and coaching industries.

Back in 2018, Amazon tried to create an engine to screen potential job applicants and create a list of the top five candidates². However, the engineers quickly realized that the AI automatically considered male candidates to be superior. The problem? They benchmarked their data against their current engineering team, who were predominantly male.

In another 1913 case, IBM teamed up with the University of Texas to build an AI-based oncology advisor³. However, they discovered a few years later that the chatbot was giving inaccurate cancer advice. They discovered that the software was trained on a very small number of hypothetical cancer patients instead of real patient data.

Of course, AI technology has come a long way since these mishaps. However, such instances still influence public perception and illustrate the importance of training AI models on valid neural models.

How Fingerprint for Success approaches AI ethics

Ethical AI doesn't happen by accident. At Fingerprint for Success, we have implemented a specific action plan to avoid some of the common ethical risks of AI. Here are the key ethical dimensions we strive for at F4S:

A human-centric approach to AI

Many people see the 'AI debate' as one between human intelligence and artificial intelligence. However, at F4S, we firmly believe it's not possible to have the latter without the former. This is why our AI coach Marlee has been engineered based on the extensive knowledge and experience of our founder Michelle Duval. By combining Michelle's 25 years of experience as a coach with the latest behavioral data science, Coach Marlee is able to give extremely personalized and human-like advice.

At F4S, we use ChatGPT and its language model GPT3 as the foundation for Coach Marlee. This allows it to generate conversational responses rather than follow a script. These machine- learning models also enable intent recognition, so that Coach Marlee can understand what coachees need and give tailored responses. Using sentiment analysis, it can also identify emotions and respond appropriately. With a much broader range of inputs than your average chatbot, Coach Marlee is also able to practice active listening. By listening for important phrases, it can steer the conversation in the right direction and 'dig deeper' during sessions.

All of these intelligent systems go a long way towards creating an experience that feels like you're speaking to a human.

Transparent, explainable, fair, and unbiased

Putting a human in the loop helps to mitigate some of the risks of generative AI. Much of the concern around unethical and discriminatory AI is due to the training data it is modeled on. As Coach Marlee has been modeled on Michelle's wealth of experience in working with a broad range of clients, it's intentionally designed to respect and nurture all people. However, every human has blind spots, and Coach Marlee has also been supplemented with data from scientific papers to further eliminate the risk of bias.

The use of Large Language Models (LLMs) in intelligent machines is another potential cause of discrimination in AI. These deep learning models can occasionally 'hallucinate' and provide inaccurate or irrelevant responses. Like ChatGPT, we do use LLMs, however we do not solely rely on them. Instead, we can use the data gathered from Marlee's past coaching sessions (there are 50,000 hours and counting!) to provide more effective coaching .

We pride ourselves on being open and transparent about the artificial intelligence technology we use, and our friendly team is always happy to answer any questions.

Private and confidential

One common concern around AI coaching is data security and confidentiality. However, with the right technology in place, it can actually be safer than working with a human coach. While coaches may have contracts in place with their clients, it's impossible to monitor or control the information they share outside their sessions.

With an AI coaching platform, this is much more predictable. As we understand how crucial trust is in the coaching process, we baked rigorous security into our platform from day one. You can rest assured that all information shared in your coaching sessions stays safe with Coach Marlee. We prioritize confidentiality by anonymizing any identifiable information and generalizing the concepts discussed. This enables Coach Marlee to learn from our sessions and provide more effective coaching without compromising user data.

These strong data privacy measures apply to all other aspects of the F4S experience too, including our talent acquisition and team collaboration features. If you have any other concerns about how we address security at F4S, please read our privacy policy before getting started.

Safe, secure, and sustainable

All of the aforementioned engineering ethics go a long way to creating F4S's safe and secure platform. And we intend to keep it that way. Where many AI companies run into trouble is trying to do too much, too quickly. They become over-excited about innovative new technology, and race it out the door before they're actually ready.

At F4S, we don't believe in disruption for the sake of disruption — especially at the expense of safety. Instead, we're committed to continuous iteration, to better serve the needs of our users. F4S has been around since 2016, making us the world's first AI coach. Since then, we have been continuously improving the Coach Marlee experience, and thoughtfully releasing new features only as needed.

What are the benefits of ethical AI for coaching?

With the right guidelines in place, AI technology can not only preserve the safety and efficiency of coaching — but actually improve it. These are some of the key benefits of ethical AI coaching.

More consistency and transparency in coaching

The coaching industry has also exploded in recent years, making it the second fast-growing sector after IT⁴. As opposed to becoming a therapist or a psychologist, you do not require any specific training or certification to call yourself a coach. While there are plenty of excellent, highly trained coaches out there, there are also plenty who lack experience in the area they are coaching in. As a result, there's limited transparency or control over the quality of coaching provided. AI coaches are trained on only the most accurate data, providing a reliable and consistent experience every time.

Greater diversity and inclusion

Every coach has their set of moral guidelines they live and work by and subsequent moral decisions they must make. These are formed largely through the lens of their own personal experience. This means that even with the best of intentions, the advice they give their clients may be muddied by their own personal biases and assumptions. For example, a coach who is Caucasian may not understand the unique barriers that people of color can experience in the workplace. On the other hand, AI coaching can draw upon a much wider breadth of information to provide appropriate feedback.

Improved accessibility

Many of the people who could benefit most from coaching are unable to access it, due to prohibitive pricing or location restrictions. For example, India is currently experiencing a mental health epidemic, due to a shortage of doctors and psychologists⁵. In the aftermath of COVID, people require more mental health support than ever. However, there are not enough specialists to treat it — and many people are unable to afford it. While AI coaching can not completely replace mental health professionals, it can help bridge the gap by making support accessible.

Accelerated progress

There are many organizations tackling some of the world's most challenging problems — whether it's a green tech start-up tackling climate change, or a not-for-profit solving world hunger. Traditionally, these companies are under-resourced, compared to those in the private sector. AI coaching allows professional development to be rolled out at scale in these companies. As a result, employees at all levels are equipped with the tools they need to reach their full potential in the workplace. Through virtual wellbeing coaching, they're also empowered to look after their own physical and mental health. As a result, they're empowered to solve these pressing social issues more quickly.

Will AI replace human coaches?

One of the common ethical dilemmas we come up against at F4S is the question, 'Won't AI coaching mean there are fewer jobs for coaches?' It is certainly true that AI technology will replace some jobs across all industries, and that includes coaching.

However, from our perspective, it also allows coaching to reach a much wider audience than ever before. Due to the high cost of human coaching, it has typically been a luxury reserved for only C-suite executives. While everyone can benefit from coaching, most people can't afford it out of their own pocket. AI coaching solves this problem, by making it accessible to people from all walks of life.

We understand that there will always be people who prefer to work with humans and that's absolutely fine. At F4S, we also work with some of the world's best human coaches, who are specifically trained in our coaching method. However, we find that supplementing this approach with sessions with Coach Marlee leads to even better outcomes — and many people quickly embrace the idea of AI coaching once they've tried it.

The F4S AI code of ethics

At F4S, our approach to ethical AI is informed by our broader values. These include:

• Make it happen: We make human growth and development accessible to all because everyone deserves the opportunity to make their goals a reality.
• Do good: By using AI thoughtfully and responsibly, we leverage its power to create positive change in the world.
• Celebrate uniqueness: We help individuals and teams embrace diversity to work better together and create an amazing company culture.
• Be curious: We listen intently to the needs of our customers, and this is reflected in the way Coach Marlee communicates with users, too.
• Delight: We expect each and every client to be pleasantly surprised by their experience of working with an AI coach for the first time.

We have a designated ethics committee at Fingerprint for Success, which helps ensure these guidelines are embedded into everything we do.

The future of ethical AI for F4S

There's no doubt that artificial intelligence technology is here to stay — in fact, it will only become more entwined in our daily lives in years to come. In order to adapt to the future of work, organizations must learn to work with intelligent systems, not against them. This means having an executable plan to integrate it into their operations.

At Fingerprint for Success we are proudly leading this shift by thinking outside the box of traditional 1-1 coaching. Our integrations with Google Calendar (coming soon) and Zoom are examples of how AI can improve the daily lives of our clients by allowing them to communicate more effectively. This is only the beginning of our quest to combine human intelligence with AI to change the world.