Artificial Intelligence (AI) is quickly transforming industries as we know them, bringing a multitude of new possibilities. But unfortunately, it’s not all sunshine and rainbows. Leveraging AI understandably comes with significant complexities and challenges, so how can we mitigate the risks?  

First and foremost, why do we trust AI? 

Trust in AI stems from several factors. Firstly, security technology ensures robust protection for data and algorithms. Companies with strong reputations prioritise safe AI practices, demonstrating a care for their reputation. Additionally, incorporating ethical values into AI development fosters trust, and adhering to laws and regulations ensures compliance and further builds confidence in AI systems. 

Can AI Regulation Keep Us Safe? 

Different regions have varying approaches to AI regulation and laws. Looking at the statistics, China implemented regulations on recommendation algorithms in 2021, soon followed by rules for deep synthesis in 2022. Meanwhile, the European Union has made efforts to introduce an EU AI Act in 2021, which is expected to be revised and expanded by 2026. In the USA, the Algorithmic Accountability Act was enacted in 2022, alongside multiple state-level data privacy laws. On the contrary, Switzerland still has some catching up to do and is still in the process of analysing potential AI regulations. They have significant efforts underway for 2024, so how can Swiss businesses get on board? 

Recommended Actions for Swiss Companies 

In preparation for the EU AI Act, Swiss companies should take various key actions. They need to create transparency by clearly communicating AI use, track use cases by keeping detailed records of AI applications, and avoid high-risk applications, especially as providers. It is also important to understand and fulfill obligations by staying informed and compliant with legal requirements. 

Suggestions to incorporate trustworthy AI 

To build and ensure trust in AI, companies could adopt a four-step approach. The first stage begins with translating ethical guidelines into practical actions, followed by integration of these actions. Next, regularly calibrating the systems to maintain ethical standards and lastly, proliferate these practices across all AI applications. 

Key Takeaways from Swiss Post & die Mobiliar 

While compliance is crucial, it alone isn’t enough to build trust in AI. Instead, it’s vital to learn from AI failures to avoid repeating the same mistakes. Furthermore, digital ethics are essential for building and maintaining trust in AI. The benefits of AI are immense, but addressing its risks is crucial for responsible and ethical development. Let’s work together to keep AI on track.