Anand Tamboli, the Author of the upcoming book on “Keeping Your AI Under Control: A Pragmatic Guide to Identifying, Evaluating and Quantifying Risks” participates in Risk Roundup to discuss Responsible AI
Responsible AI
AI technologies bring the transformative power to nations: their government, industries, organizations, and academia (NGIOA). The development of AI is creating new opportunities for everyone, even individuals. As AI technologies become more pervasive and get deeply embedded in products and services, and responsible for an increasing number of decision-making processes like benefit payments, mortgage approvals, parole grants, college admissions, job interview screening, and medical diagnosis, they become less visible and transparent.
Since algorithms are not viewable, one of the real risks with AI is amplifying and reinforcing existing human biases into AI decision-making processes. While the biases could be either intended or unintended, the reality is that AI-based decisions should be understandable and auditable to those impacted and adhere to existing rules and regulations.
While the goal of emerging technologies like AI is to improve the lives of everyone around the world, it is also raising further questions about the best way to build fairness, interpretability, responsibility, accountability, privacy, and security into these emerging systems. These issues are far from solved, and in fact, at the forefront of the AI technology adoption across nations.
Even though AI is quickly becoming a new tool for transformation, it has also become clear that deploying AI requires careful management and governance to prevent unintentional damage to individuals and society as a whole. Justifiably, trust in decision-making AI systems is going to be crucial as we move forward in using AI systems for decision-making broadly. Perhaps coding responsibility, accountability, and explainability into algorithms will be our only solution.
Understanding Responsible AI
Responsible AI is about building trust in AI solutions and currently focuses on ensuring the ethical, transparent, and accountable use of AI technologies in a manner consistent with user expectations, organizational values, and societal laws and norms. The question is whether this is enough and effective.
The goal is to have a Responsible AI guard against the use of biased data or algorithms, ensure that automated decisions are justified and explainable, and help maintain user trust, individual privacy, and security. While there is a broad hope that it can be made possible by providing clear rules of engagement, the question is unless the rules are coded, ensuring the responsibility, accountability, and explainability will perhaps not be possible. Responsible AI that is in the code will allow organizations to innovate and realize the transformative potential of AI that is both compelling and accountable and make it easier for explainable AI to be accountable, explainable, and effective.
While at the moment Responsible AI is about creating governance frameworks to evaluate, deploy, and monitor AI to create new opportunities, it requires architecting responsibility in the code and implementing coded solutions that put humans at the center. By using design-led thinking, organizations at all levels can examine core ethical questions in context right in the code, evaluate the adequacy of policies and programs, and create a set of value-driven requirements to govern AI solutions. That brings us an important issue, how should automated decision systems be governed? Simply by governance frameworks or using code as the constitution for embedded responsibility, accountability, and explainability.
In the coming years, responsible AI will need to be a critical component of algorithms as well as an organizational change model that focuses on rapid learning and adapting. It is time to define a framework for how responsible AI can be embedded in the code, and security checkpoints are assigned to create checks and balances for this process. By integrating responsible AI into the system and organizational approach for change, it is possible to ensure that the critical element of trust is cultivated and maintained among critical human stakeholders, the most important of which being employees, customers, citizens, and consumers.
The time is now to define a practical Responsible AI framework that will certainly enhance Explainable AI and accountability further.
For more, please watch the Risk Roundup Webcast or hear the Risk Roundup Podcast
About the Guest
Anand Tamboli is the author of the upcoming book on “Keeping Your AI Under Control: A Pragmatic Guide to Identifying, Evaluating and Quantifying Risks”.
About the Host of Risk Roundup
Jayshree Pandya (née Bhatt), the founder and chief executive officer of Risk Group LLC(www.riskgroupllc.com), is working passionately to define a new security-centric operating system for humanity. Her efforts towards building a strategic security risk analytics platform are to equip the global strategic security community with the tools and culture to collectively imagine the strategic security risks to our future and to define and design a new security-centric operating system for the future of humanity.
About Risk Roundup
Risk Roundup, a global initiative launched by Risk Group, is a security risk reporting for risks emerging from existing and emerging technologies, technology convergence, and transformation happening across cyberspace, aquaspace, geospace, and space. Risk Roundup is released in both audio (Podcast) and video (Webcast) format and is available for subscription at (Risk Group Website, iTunes, Google Play, Stitcher Radio, Android, and Risk Group Professional Social Media).
About Risk Group
Risk Group LLC is a leading strategic security risk analytics platform.
Copyright Risk Group LLC. All Rights Reserved
Emerging Technology Trends 