Sonova’s commitment to digital ethics in AI

Artificial intelligence (AI) is rapidly becoming a focal point across technological fields, marking a pivotal era of innovation. Since 2000, AI has driven numerous advancements at Phonak, and today, we believe we are entering an even more transformative phase.

AI is set to reshape the delivery of hearing care services and revolutionize the functionality of hearing aids, ultimately enhancing the way patients experience improved hearing and other associated health and well-being benefits.

Sonova’s digital ethics commitments

At Sonova, we recognize that to fully leverage AI’s potential, we must stay committed to the ethical use of data. By combining innovation with ethical guidelines, we aim to ensure that our technological advancements benefit society while upholding fundamental values.

Since 2022, Sonova’s Compliance team has been continuously developing a comprehensive digital ethics and AI governance framework. This framework ensures consistent and responsible use of AI in hearing devices and internal tools that improve productivity.

AI relies on large amounts of data and powerful processing to train its models. We recognize that this data can have inherent biases (e.g., ethnic backgrounds, ability levels). To address this, we continually work to refine our models to minimize bias and ensure more equitable outcomes for all users.

Sonova’s digital ethics framework requires appropriate assessments before an application with AI is deployed, ensuring responsible use. This includes:

1. Privacy Risk Assessments: The Compliance team uses privacy risk assessments like the Data Protection Impact Assessment (DPIA) to evaluate the risks of using AI, especially in collecting and processing personal data. If AI is involved, an AI Threshold Risk Assessment is also required.

2. AI Threshold Risk Assessment: Based on the DPIA findings, we implement further measures to prevent rights violations and avoid non-compliance with laws, such as the EU AI Act¹, which could lead to significant fines. The AI Threshold Risk Assessment helps us understand the legal obligations related to identified risks, including transparency requirements and the need for fundamental rights impact assessments.

3. Digital Ethics Working Group: Sonova has set up a digital ethics working group made up of internal experts and stakeholders. This group discusses AI developments and ensures new AI products and services comply with regulations.

The European Union’s AI Act and Ethical AI Use

In line with its commitment to ethical innovation, the European Union has introduced the AI Act, a comprehensive regulation to assess and manage AI risks to citizens.¹ The AI Act uses a risk-based approach and classifies medical devices, including hearing aids, as “high risk.” This classification requires strict assessments and compliance measures.

The EU’s regulatory efforts, along with proactive steps by companies like Sonova, show a growing awareness of the importance of ethics in the digital age. As we move forward with new technologies, keeping ethics in focus will be key to building a fair and prosperous future.

We invite you to visit our website to learn about the latest AI advancements in Phonak Infinio hearing aids.

---

References:

1. The EU Artificial Intelligence Act. Retrieved from The EU Artificial Intelligence Act (artificial-intelligence-act.com), accessed September 2024.