Unlocking ethical AI: Fostering transparency and collaboration in organizational transformation

Jaideep Kewalramani,  Head of Employability Business & COO, TeamLease Edtech recently spoke to Dataquest. Jaideep has over 25 years of experience in leading and developing businesses across technology, outsourcing, digital and AI. He also holds 4 patents in the Artificial Intelligence (AI) domain and has been recognized as an ecosystem contributor by IndiaAI, which is a Goverment of India Initiative). He is also the founding member of the Unmanned & Autonomous Vehicles Association of India.

DQ: Where should Data and AI figure in the business growth objectives of organisations?

Jaideep Kewalramani: Since the rise of generative AI, its significance has pervaded every echelon of organizations, from the boardroom to the most junior roles, whether its application is formal or informal. However, the acknowledgment AI deserves has not been universal in addressing your query. To be more specific, it should be among the top three priorities for the C-suite in any organization. I have persistently advocated for the appointment of a Chief AI Officer, a role distinct from the CTO or CIO. Initially, this could be a dual role, but given the rapid pace of technological advancements and the emergence of real-world use cases, all organizations will inevitably find themselves crafting AI policies, strategies, and exploring diverse use cases. Consequently, the necessity for a dedicated Chief AI Officer will become increasingly evident, standing as one of the top three priorities and demanding substantial focus.

DQ: How can organizations effectively connect their data and AI initiatives with overall business growth strategies? 

Jaideep Kewalramani: At the strategic level, channeling your data through AI engines can yield a plethora of insights that often elude typical tools or human analysis, thereby fostering a virtuous cycle of uncovering diverse patterns and data nuances. This phenomenon operates on a strategic plane. On a tactical front, data ought to undergo thorough AI scrutiny to sift out noise and inefficiencies. The goal is to procure pristine data, achievable either during input, processing, or even output stages. At the transactional level, employing AI to interact with data collection becomes paramount, facilitating the prediction of patterns and ensuring a seamlessly enhanced user experience. As teams navigate data utilization, AI intervention can aid in prioritization and furnish real-time insights tailored to their specific needs, culminating in heightened predictability. This synergy between data and AI operates across all three levels. Hence, it becomes imperative for organizations to recognize and harness the pivotal amalgamation of data and AI.

DQ: What are the common challenges organizations face when trying to integrate AI and data with furthering the growth of their organisation? 

Jaideep Kewalramani: The initial challenge that every organization encounters involves identifying tools that may present themselves as AI but lack authenticity. It's crucial for organizations to exercise discernment and apply filters to avoid mistaking intelligent data management systems for genuine AI solutions. This stands as the first pitfall they should steer clear of. The other two critical areas that demand an organization's attention while constructing their data and AI strategy are twofold. Firstly, they should assess whether investing in AI will yield a tangible and effective return on investment at present. Jumping onto the AI bandwagon simply for trendiness should be avoided; instead, focus should be on clearly defining use cases and collaborating with partners to gauge how AI will enhance end-client interests and user experiences. Selecting pilot projects for scalability evaluation comes next. Lastly, equally vital is addressing data security and privacy concerns inherent in AI tool implementation. With data often drawn into the cloud, questions about cloud ownership, jurisdiction, and security arise. Navigating these aspects is of utmost importance. As the array of applications continues to expand, careful consideration is paramount in choosing partners and approaches.

DQ: What role does data governance play in connecting data and AI with business growth, and what challenges do organizations encounter in establishing effective governance frameworks? 

Jaideep Kewalramani: The role of AI is undeniably prominent, and it can be conceptualized quite powerfully by likening AI tools to external human interactions. The simplicity lies in a question: if you're comfortable sharing specific data with someone outside the organization, the same comfort should extend to sharing that data with AI tools. This serves as a straightforward yet practical filter applicable to any C-suite executive. Governance structures and standards also play a vital role. Different layers of the organization require distinct governance standards. At the grassroots, where human-AI interactions occur or AI-augmented humans are involved, there should exist a clear organizational policy delineating acceptable data input boundaries. The definition of use cases and establishment of reporting mechanisms should follow, encouraging AI utilization while obtaining permissions for specific AI tool applications through a streamlined review process. These protocols ensure safeguards are in place.

When considering proprietary or in-house AI tools using organizational data, challenges are minimal as data remains internal. However, when introducing third-party AI tools, the same scrutiny applied to procuring cloud or blockchain solutions must be enacted. Identical diligence should be upheld for data security when collaborating with third-party applications. Just as with acquiring a new technology solution, the evaluation process remains essential in ensuring the preservation of data integrity and security.

DQ: What are the key considerations for organizations when selecting and implementing AI technologies to support their business?

Jaideep Kewalramani: Let's delve into each of these components. First, let's address the contentious issue at hand. Personally, I don't advocate for the complete banning of AI, especially generative AI. The reality is that there will always be both positive and negative applications, with some using AI to advance society and others to cause harm. However, if those with positive intentions halt AI development, it merely gives a head start to those with malicious motives to refine their tactics. In today's rapid-paced landscape, even a short 6-month pause could result in a 5-year setback. This emphasizes the urgency of our situation.

Now, moving on to the other points you raised about utilizing AI profitably while maintaining ethical standards. Ethical AI usage is heavily influenced by organizational culture and policies. Organizations must foster a climate that encourages both AI adoption and the transparent sharing of AI usage experiences. Those who employ AI tools in their workflow should be supported in doing so openly. Managers can play a crucial role by recognizing and rewarding such initiatives. Moreover, this practice opens the door to discussions about purchasing policies and data security, ensuring ethical considerations are met.

Rather than stifling AI usage, fostering an environment of open disclosure is key. If an organization expresses a desire to become AI-ready, individuals within the organization should be welcomed to become AI champions. Running pilot projects can help cultivate a culture of inclusivity and transparency, preventing unintended ethical breaches that may occur when employees resort to using tools without organizational awareness. Full disclosure, along with respecting intellectual property, becomes integral.

For instance, when employing generative AI for commercial content, proper credit attribution to the source is essential, and intellectual property boundaries must be upheld. It's also imperative not to blindly trust generative AI models, as they can have inherent issues like hallucinations. Organizations must exercise caution when incorporating AI-generated content into their products and prioritize human review.

Lastly, I'd like to emphasize the significance of organizations embracing AI-augmented humans. Beyond the Chief AI Officer, the Chief Human Resources Officer (CHRO) has a pivotal role in integrating AI to enhance human capabilities, rather than replacing them. This shift in approach has the potential to be a transformative game-changer for many organizations.