Menu Back to Poster-Presentations-Details

P228: Development of Data and Artificial Intelligence (AI) Ethics Principles within a Pharmaceutical and Diagnostics Company

Poster Presenter

Timothe Menard

Global Head, Quality Excellence Digital | Bioethics Coach (Data Ethics)
F. Hoffmann-La Roche Ltd
Switzerland

Objectives

The primary goal of this initiative was to develop a set of robust Data and AI Ethics Principles within a pharmaceutical/diagnostics company. These principles aim to create a comprehensive ethical framework that aligns with global standards and the specific processes of the company.

Method

After benchmarking available principles across pharma and professional societies, and peer-reviewed journal publications; data and AI Ethics principles were collaboratively drafted by the internal Bioethics department, policy leads, and experts in AI, statistics, data privacy, and related fields.

Results

Data Ethics Principles: Beneficence: Our use of data maximizes social good Autonomy: Respect individuals consent, privacy & protect their rights Transparency: Individuals informed on how their personal data are used Data Quality: Fit for purpose, high quality data used to make decisions Data Integrity: Optimize insights by promoting data integrity Fairness, Non-Discrimination, Justice: Data should be equitable, inclusive Ethics by Design: Controls to prevent harm & risks to individuals are built into the design of data architecture/processing Responsible Data Sharing: Data sharing based on processes that actively & consistently consider, prioritize and protect individuals rights Accountability: Monitoring, governance & disciplinary actions operationalized Proportionality: Data collected with due consideration for risks/benefits, data used in a purposeful manner Privacy: Respect privacy of all individuals, act in accordance with relevant laws & regulations Security: Data systems designed to be secure Sustainability: Data processed efficiently to ensure small carbon footprint AI Ethics Principles: Ethical Use: Maximizing benefit of AI & minimizing harm Transparency: Transparent about use of AI to build trust & credibility with patients, and society at large Explainability: Ability of humans to understand & interpret AI solutions Human Control: Humans can override decisions made by AI systems & machine autonomy can be restricted Empowering People: Enhancing shared decision-making of patients, carers and HCPs Accountability: Monitoring, feedback, governance actions are operationalized Fairness, Minimization of Bias: AI tools should be inclusive, equitable, and seek to support the mission of responding to the needs of all patients Privacy: Data for AI models are privacy-protected Security: AI systems designed to be secure Safety by Design: Safety embedded in the design of AI systems Sustainability: Think about the future impact of AI system

Conclusion

The development and implementation of robust Data and AI Ethics Principles represent a key step in ensuring ethical considerations are at the core of our decision-making processes. The Data Ethics principles underscore the commitment to ethical data management. They prioritize ethical behavior while supporting the mission to meet the needs of patients, as we recognize there is a person behind every data point. Ethical behavior is critical across the data lifecycle, from collection, use, overall management and storage across the development and commercialization of our products and technologies. The AI Ethics Principles highlight our dedication to the ethical use of AI and the future impact of AI systems on society and the planet, with a critical eye to the challenges of AI. Both sets of ethical principles are written in lay language so that the general public (non-scientists) are able to understand them. We established a communication plan to disseminate these principles both within and outside the organization. Interactive training materials were designed for all employees, including a specific module with advanced applied content for data practitioners. These principles align with global ethical standards while also reflecting on our organization's unique contextual requirements and our corporate values (integrity, courage, passion). Our journey underscores the importance of proactively addressing ethical concerns in AI and data applications. This is crucial not only for organizational integrity but also in service to global society. By following these principles, our organization aims to be prepared to meet the evolving challenges and opportunities of the AI and data-driven future while maintaining a strong commitment to ethical practices.