Bias in AI are at the heart of concerns: Capgemini VP

0


UNESCO’s framework for ethical AI would have a dramatic impact on all AI activities, including development, enforcement, ethics, data privacy and regulation, according to Padmashree shagrithaya, VP, Analytics & Artificial Intelligence – India, Capgemini in an interview. Excerpts:

What are the challenges of creating impartial, non-discriminatory and sensitive AI platforms?

AI systems, we know, are built for the purpose of mimicking human intelligence. Systems are taught by sharing huge data about past human actions to learn from. While this is useful on the one hand, to reduce human interference in repetitive decision-making and enable the resolution of larger and more complex problems, this approach is forged with inherent challenges of prejudice and discrimination. Historical data is littered with prejudices and discriminations that are sometimes deliberate and often unintentional. How to identify them, separate them / correct them and send them back to the algorithm to ignore them or systematically adjust them for the same, is one of the biggest challenges.

Is forcing consumer / patient / citizen consent through pop-ups enough to move on?

The general opinion is that the more information we have about individuals, the more we could create more precise algorithms to target them for offers / recommendations / treatments etc. There is a lot of discussion around this, for the breach of personal and sensitive information and its negative impact. They are useful in many scenarios, such as recommending the right treatment plan to patients or providing helpful recommendations that consumers are looking for. But the question again is where to draw the line on how much information about an individual would amount to infringing on sensitive personal data.

What are the main things AI developers should keep in mind to avoid gender bias?

Bias in AI is a key concern and there is a lot of research going on around it. In fact, gender bias has been a focal point. It’s important to address this systematically throughout the AI ​​development lifecycle, starting with scoping through to proof that it’s been supported. Developers should ensure that the scope is neither nearsighted nor intended to introduce an intersection mismatch. Plus, keeping in mind the overall societal impact that the results can have. Data collection and retention and the methodology involved will play a very important role in ensuring fairness and the ethical end result. Accountability and ownership should be well defined, to ensure fairness and inclusion throughout the development, validation and delivery of an AI system. Most importantly, all of these must be part of a regulatory and compliance framework.

Are technology developers and deployers sufficiently sensitive to these critical issues?

There is still a great degree of ignorance about how AI can impact our lives. In recent years, we have seen rigorous campaigns to raise awareness of sensitivity issues and also to bring about changes in the mindset of leaders. But still, a large ground must be covered. In this journey, awareness is not enough, but recognition, acceptance and dealing with the problem is more important. We should have legal frameworks that support holistic inclusion of coders, developers and decision makers. Data collection approaches and data distribution should be streamlined to avoid bias against under-represented groups (if any). Most importantly, a collective effort of the AI ​​community is needed to create a sustainable ecosystem to ensure transparency and gain the trust of users, which is of paramount importance for any system focused on technology works efficiently.

UNESCO is working on a framework to ensure ethical AI … what does this mean for AI developers and users?

UNESCO’s framework for ethical AI will have a far-reaching impact in the AI ​​space. The idea is to have a holistic and evolving framework of values, principles and actions that can guide companies to deal responsibly, with the known and unknown impact of AI on humans and society. in general. Through this, UNESCO aims to secure a global commitment from each country to view AI from an ethical perspective. The framework will have impacts on a wide range of areas such as privacy awareness and inclusion; transparency, fairness and non-discrimination; empowerment through participation; a change of mindset that supports a sustainable AI environment and a fair balance between business growth and the promotion of human values.

What is Capgemini’s contribution to building an ethical and fair AI environment?

At Capgemini, we believe ethical AI is the cornerstone on which customer trust and loyalty is built. From conceptualization to development, delivery and monitoring of AI systems, we take great care to ensure that we are building a trustworthy AI solution. Our Code of Ethics for AI has seven key principles that form the building blocks of our AI systems. We have established an “Ethics AI Review Committee” to review AI projects and report independently on practices adopted and adherence to the principles. We have our own solution framework (AI Glass Box and AI Fairness Tool-fAIry) allowing companies to visualize complex topics such as transparency, bias and fairness with simplicity and also for customers to identify and measure the equity of an AI model, respectively.

What kind of partnership with stakeholders is needed to make search engine technology more secular, neutral and unbiased?

Search engines have the power to shape user behavior. While the entire cycle of collecting, indexing and ranking content is automated, it is still subject to abuse and unintentional bias. For search engines to be secular, neutral, impartial and conflict-free, their searches must be objective, based on the continuous learning of algorithms, and they must be smart enough to deliver the right search results. Many ambiguities can be resolved to secure human values ​​and societal benefits through partnership with stakeholders involving AI vendors, developers, end users, regulators, compliance agencies, industry experts and search engine providers.


Share.

About Author

Comments are closed.