Share this post:
Few technologies attract the level of attention that artificial intelligence (AI) does from governments, business, academics, media and the public, especially at a relatively early stage in their adoption. This attention reflects both the enormous positive potential that all stakeholders recognize in AI, as well as the many concerns – some well-founded, others perhaps less so.
The European Commission has been at the forefront of global efforts to understand and assess the risks and benefits of AI, and to establish ethical principles for its use. As a member of the Commission’s High-Level Expert Group on AI, IBM has helped developEurope’s ethical principles on AI last year, and welcomed similar work from the OECD. Such principles are important to help communicate commitments to citizens and consumers, and to set a direction in a complex and evolving area. However, it is now time to move from principles to clear policies that all stakeholders can rely on.
That is why IBM welcomed the opportunity to contribute to the European Commission’s consultation on their February 2020 White Paper on Artificial Intelligence. We support targeted policies that would increase the responsibilities for companies to develop and operate trustworthy AI. As the Commission White Paper recognizes, an effective governance framework must first of all be risk-based, and it must seek to regulate not the technology itself but rather its uses. We strongly agree with this overall approach, and believe it is consistent with what we set out in our January 2020 policy paper on precision regulation of AI, where we proposed that a risk-based governance framework for AI should be based on the pillars of accountability, transparency, fairness and security.
You can download our detailed views on the Commission White Paper below, but we want to highlight some of our key points on the proposed regulatory framework for AI here:
- For regulatory purposes the definition of AI systems needs to be narrow and clear, so that the focus is on those systems that cause serious concern and avoiding non-AI systems being inadvertently included.
- There should be a single risk assessment framework to identify high-risk applications regardless of sector, and without lists of exceptions.
- In any future regulatory framework, legal obligations should be imposed on the actor(s) in the AI supply chain best placed to address any potential risks.
- In sectors where established conformity assessment mechanisms already exist, these should also cover high-risk AI applications in those sectors. In other sectors, an appropriate level of compliance for high-risk applications can be achieved with a combination of ex-ante self-assessment and ex-post auditing and enforcement.
- While voluntary labelling systems can be helpful to consumers or end-users in some markets, we do not believe they would be effective across such a broad field as AI applications.
- We believe facial recognition is one of the most concerning applications of AI. That is why our CEO Arvind Krishna has recently communicated that IBM no longer offers general purpose IBM facial recognition or analysis software. IBM firmly opposes and will not condone uses of any technology, including facial recognition technology offered by other vendors, for mass surveillance, racial profiling, violations of basic human rights and freedoms, or any purpose which is not consistent with our values and Principles for Trust and Transparency.
As the discussion on Europe’s regulatory approach continues, IBM will continue to advocate for it to be in line with our Principles for Trust and Transparency:
- The purpose of AI is to augment – not replace – human intelligence.
- Data and insights belong to their creator.
- New technology, including AI systems, must be transparent and explainable.
Now is the time to turn these principles into real policies.
— Barry O’Brien, Government and Regulatory Affairs Executive, Europe
— Jean-Marc Leclerc, Head of EU Government and Regulatory Affairs
For more information and IBM’s full response to the European Commission’s consultation on the White Paper on AI, click to download the PDF: