{"id":701009,"date":"2025-07-15T12:50:11","date_gmt":"2025-07-15T10:50:11","guid":{"rendered":"https:\/\/www.devoteam.com\/expert-view\/the-future-of-responsible-ai-5-aspects-we-should-talk-about\/"},"modified":"2025-09-01T12:34:38","modified_gmt":"2025-09-01T10:34:38","slug":"the-future-of-responsible-ai-5-aspects-we-should-talk-about","status":"publish","type":"expert-view","link":"https:\/\/devoteam.info\/en-nl\/expert-view\/the-future-of-responsible-ai-5-aspects-we-should-talk-about\/","title":{"rendered":"The Future of Responsible AI: 5 Aspects we Should Talk About"},"content":{"rendered":"\n

<\/path><\/svg><\/span><\/span>Estimated reading time: <\/span>10<\/span> minutes<\/span><\/p>\n\n

The idea of giving rules to intelligent machines is not new.<\/span><\/strong> Back in 1942, science fiction writer Isaac Asimov<\/span>\u00a0introduced his famous\u00a0“Three Laws of Robotics”<\/strong><\/span>. These fictional laws were part of the stories collected in \u201cI, Robot<\/em>\u201d and they were designed to ensure robots served humanity safely.\u00a0<\/p>\n\n

While coming from science fiction, Asimov’s work sparked an important thought relevant today: what rules should<\/em> we give the intelligent systems we’re building?<\/strong> As AI takes on more tasks autonomously, it prompts us to think carefully about our own values. What decisions are we comfortable letting software make for us? This is the aim of this article – to join the discussion about the future of ethical and responsible AI from the perspective of a tech expert. If you are new to the topic, I recommend starting with my article explaining the basics of ethical AI<\/a>.<\/p>\n\n

The Future of Responsible AI: 5 things to consider<\/strong><\/h2>\n\n

As an inspiration for our further exploration, let’s start with The Three Laws of Robotics<\/strong>, presented in the fictional “Handbook of Robotics, 56th Edition, 2058 A.D.”<\/p>\n\n

\"Three<\/figure>\n\n

How does it translate to the foundations of ethical AI? Firstly, avoiding causing harm is at the core of ethical AI frameworks and principles<\/a>. The second law corresponds with the need for AI to operate under human control and align with human intent, while acknowledging that this obedience must always be subservient to the avoidance of harm. Finally, the third law echoes discussions around accountability and the need for clear hierarchies of ethical priorities in autonomous systems.<\/p>\n\n

Time to delve deeper into the 5 aspects of responsible AI for a better future.<\/p>\n\n

1. The Enduring Compass: Our Core Ethics Still Point the Way<\/strong><\/h3>\n\n

The good news\u00a0is that we don’t need to invent a completely new set of morals for AI. Fundamental human ethical principles \u2014 ideas like\u00a0fairness, justice, doing good, avoiding harm, being accountable, and transparency<\/strong> \u2014<\/span> remain incredibly relevant.\u00a0<\/p>\n\n

These principles, developed over centuries of human experience and thought, provide the moral compass we need. They are based on human rights and the kind of society we want to live in. AI, ultimately, is a tool created by humans, and it should serve these enduring values. At Devoteam, we believe in \u201cTech For People\u201d, it is our ethics compass that has stood the test of time.<\/p>\n\n

\n
\"Quote\"<\/figure>\n\n\n\n

Businesses must ensure that their AI systems are fair, transparent, and accountable. They should avoid bias, use AI responsibly, protect data, and follow the rules.<\/p>\n\n\n\n

\n\n\n\n
\n
\"\"<\/figure>\n\n\n\n
\n

Gert Jan van Halem<\/strong>Chief Technology Officer at Devoteam<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n

\n

Real-life example:<\/strong> A powerful illustration of using AI in alignment with our ethics is the work of the company Zipline<\/a>. It operates the world’s largest automated, on-demand delivery service for medical supplies. In countries like Rwanda and Ghana, Zipline uses autonomous drones to deliver blood, vaccines, and other critical medical products to remote hospitals and health clinics, often in a fraction of the time it would take by road.\u00a0<\/p>\n<\/div>\n\n

2. The Translation Challenge: Teaching Machines Our Human Values<\/strong><\/h3>\n\n

This is the tricky part. We humans learn ethics through experience<\/strong>, empathy, and understanding context, often<\/strong> without explicit rules.<\/strong> AI systems don’t have consciousness, feelings, or lived experience. So, how do we translate our complex, often subtle, human values into concrete instructions that an AI can understand and follow? How do you program “compassion” or ensure “fairness”<\/strong> works correctly in every unexpected situation, for example, when an algorithm helps decide on school applications or job candidates? This is one of the central questions when talking about the future of responsible AI: how to apply our human ethical framework to non-human ways of thinking.<\/p>\n\n

\n
\"Quote\"<\/figure>\n\n\n\n

It is not intelligence that defines humanity; it is purpose, empathy, and intention. And no machine can replicate that, only imitate and reinforce it.<\/p>\n\n\n\n

\n\n\n\n
\n
\"image<\/figure>\n\n\n\n
\n

Patr\u00edcia Milheiro<\/strong>Strategy and Engagement, Devoteam AI Agency<\/p>\n<\/div>\n<\/div>\n<\/div>\n\n

\n

Real-life example:<\/strong> A well-known example of the ethical challenges of AI is COMPAS<\/a>, a tool that helps judges predict recidivism. However, the system was found to contain discriminatory bias, leading to unfair rulings based on race and other sensitive factors.\u00a0<\/p>\n<\/div>\n\n

3. Who’s in Charge? Grey Areas in Autonomy and Accountability<\/strong><\/h3>\n\n

AI is becoming increasingly autonomous. It can make decisions and take actions with less direct human input. This also generates ethical dilemmas: who is responsible if things go wrong? Is it the programmer, the owner, the person using the AI, or perhaps even the AI itself <\/strong>(which opens up big legal questions)?\u00a0<\/p>\n\n

When humans make difficult choices, they can be held accountable and asked to explain their reasoning, perhaps even in court. How do we build similar accountability into AI? Can we test AI in simulations beforehand to understand its choices in tough ethical situations? And if we ask AI to be able to explain everything – even what we as humans can\u2019t – it raises the question: Are we holding AI to an even higher standard than we hold ourselves?<\/p>\n\n

One answer to this challenge might be in\u00a0“human-in-the-loop” systems<\/a>, which combine human intelligence with machine learning capabilities. This approach involves active and continuous human participation, integrating humans into the AI process flow. The goal is to use human input to maximise AI’s potential<\/span> while mitigating its risks.<\/p>\n\n

\n

Real-life example:<\/strong> In March 2018, a self-driving test vehicle struck and killed a pedestrian<\/a> in Tempe, Arizona. The incident triggered a significant investigation and debate over accountability. The National Transportation Safety Board (NTSB) investigation revealed a complex chain of failures, raising profound legal and ethical questions about who is ultimately responsible.<\/p>\n<\/div>\n\n

4. Building Better AI: Practice, Transparency, and Working Together<\/strong><\/h3>\n\n

Applying ethics to AI isn’t just theory<\/strong>; it requires practical steps. Developers and researchers are working on ways to do this, such as:<\/p>\n\n