Neglecting AI Ethics Poses Huge Risks: How to Implement AI Responsibly

Back in July 1945, MH Hasham Premji laid the foundation for what would become a tech giant by starting Western India Vegetable Products Limited in the small town of Amalner, Maharashtra, India. Nestled by the Bori River, the company initially focused on producing cooking oils. Fast forward to the 1970s, and the business transformed dramatically, shifting its focus to IT and rebranding as Wipro. Today, Wipro stands as one of India's premier tech companies, operating in 167 countries, employing nearly a quarter of a million people, and generating revenue exceeding $10 billion. At the helm is Rishad Premji, the grandson of the visionary founder.

In the present day, Wipro positions itself as a global leader in IT transformation, consulting, and business process services. In an exclusive interview with ZDNET, we had the chance to sit down with Kiran Minnasandram, the VP and CTO of Wipro FullStride Cloud, to delve into the company's latest endeavors and the future of IT.

Kiran is at the forefront of Wipro's technological innovation, spearheading initiatives that develop cutting-edge solutions. His primary focus is on driving innovation and empowering businesses with state-of-the-art technology, particularly in cloud computing. He designs and implements advanced cloud architectures that revolutionize business operations, enhancing efficiency, scalability, and flexibility to help clients advance in their digital journeys.

Company Values

ZDNET: How do you define ethical AI, and why is it critical for businesses today?

Kiran Minnasandram: Ethical AI isn't just about ticking legal boxes; it's deeply connected to Wipro's core values. We anchor our approach in four key pillars: respect for the individual, fairness and transparency for society, environmental stewardship, and technical robustness. This ensures our AI aligns with our values on privacy, dignity, societal fairness, and environmental responsibility, while also being safe and compliant.

ZDNET: Why do many businesses struggle with AI ethics, and what are the key risks they should address?

KM: The challenge often stems from a lack of common language around AI. To tackle this, companies need to establish a cross-organizational strategy that integrates technical, legal, and HR teams. AI isn't just a tech issue; it's a corporate one. Businesses must define their AI ethics beyond legal compliance, focusing on the values they wish to uphold. They should also develop a risk taxonomy to address potential legal, security, and workforce impacts.

ZDNET: How does AI adoption impact corporate sustainability goals, both positively and negatively?

KM: AI has a dual impact on sustainability. On the positive side, it can improve operational efficiency, optimize supply chains, and enhance resource management, which can lead to reduced carbon emissions. For instance, in logistics, AI can optimize routes to minimize fuel use. However, the rapid deployment of AI also increases energy consumption and carbon emissions, particularly with the intensive computing power needed to train large AI models.

Environmental Impact

ZDNET: How should enterprises balance the drive for AI innovation with environmental responsibility?

KM: Enterprises need to set clear policies and guidelines for sustainable AI use. This framework helps teams make informed decisions about infrastructure and algorithms. Companies should also track and monitor AI's environmental impact, pushing service providers to do the same. We've helped clients revise their AI policies, engage stakeholders, and train employees to embed sustainability into everyday processes. Collaboration across industries is crucial for addressing these challenges, and Wipro is actively involved in global efforts to help clients navigate this complex landscape.

ZDNET: How are global regulations evolving to address ethical AI and sustainability concerns?

KM: AI doesn't operate in a vacuum. Existing laws on privacy, consumer protection, security, and human rights all apply. Data protection regulators are key in safeguarding against AI harms, while consumer protection laws address issues like algorithmic pricing. Some regions are also enacting AI-specific laws, like the EU AI Act, which imposes stricter controls based on the product's risk level. In the US, individual states are focusing on AI in labor management, one of the more complex areas of AI deployment.

Biggest Misconception

ZDNET: What are the biggest misconceptions about AI ethics and sustainability, and how can businesses overcome them?

KM: A common misconception is that innovation and responsibility can't coexist. In reality, responsible AI is crucial for sustainable innovation. Trust is the foundation of AI deployment, and companies that balance innovation with trust gain a competitive edge. Consumers and businesses alike will gravitate towards products they trust.

ZDNET: How does Wipro FullStride Cloud support companies in aligning AI with ESG (environmental, social, and governance) goals?

KM: We help by developing responsible AI frameworks that prioritize fairness, transparency, and accountability. We use AI to monitor and report ESG metrics and support Green AI initiatives to reduce the carbon footprint. On the infrastructure side, we optimize workloads for energy efficiency and develop industry-specific solutions to meet ESG targets.

ZDNET: What are the most effective ways cloud solutions can reduce AI's environmental footprint?

KM: Cloud solutions can enhance energy efficiency through renewable energy use, optimized cooling, and carbon-aware computing. Techniques like federated learning and model pruning can reduce the energy required for AI model training. Serverless and auto-scaling solutions help align resources with demand, preventing over-provisioning. Cloud providers offer carbon tracking and reporting tools, and multi-cloud and edge computing can minimize data movement, processing AI closer to its source.

Leveraging the Cloud

ZDNET: How can cloud infrastructure be leveraged to embed ethical considerations into AI development?

KM: Cloud infrastructure provides tools to embed ethical considerations into AI. AI ethics toolkits can detect bias and test for fairness, while diversity-aware training tools ensure datasets are representative. Cloud-based AI frameworks offer explainability and transparency, and secure, privacy-preserving AI development is supported through capabilities like differential privacy and encrypted processing. Automated compliance monitoring ensures adherence to regulations, and tools for model drift testing and hallucination detection help maintain model reliability over time.

ZDNET: Why do some organizations struggle to measure AI's sustainability impact, and how can cloud-based tools help?

KM: The lack of standard metrics makes it challenging to measure AI's sustainability impact. Cloud-based tools can fill this gap with customizable dashboards that track carbon output throughout the AI lifecycle. Real-time monitoring is crucial as energy consumption can vary significantly, and cloud platforms offer dynamic tracking tools to capture these fluctuations. Cloud-native solutions also aggregate data from various environments, improving visibility and decision-making. Additionally, cloud tools can highlight hidden environmental costs across the AI process and help navigate the varying regional ESG compliance requirements.

ZDNET: What concrete steps can organizations take to improve AI transparency and accountability?

KM: First, train your workforce to use AI responsibly, encouraging them to question and test AI within safe boundaries. Second, establish a comprehensive governance structure that spans all aspects of the business, from procurement to risk management.

ZDNET: How does AI bias emerge, and what role do cloud-based frameworks play in mitigating it?

KM: AI bias can arise from unrepresentative or prejudiced training data, errors in human-labeled datasets, and outdated models perpetuating historical biases. Cloud-based frameworks help by ensuring compliance with diverse regulations and validating AI models across various demographics. Adaptive training processes can also rebalance datasets to mitigate biases.

ZDNET: What governance strategies should enterprises implement to ensure responsible AI usage?

KM: A robust governance framework is essential. Some organizations might have a dedicated AI governance structure, while others, like ours, integrate it within existing governance systems. It's crucial to involve all parts of the organization and conduct AI impact assessments from the outset to embed legal protection, privacy, and robustness into AI deployment.

AI Issues

What are your thoughts on the growing emphasis on ethical and sustainable AI? Have you implemented any frameworks or policies to ensure responsible AI development in your organization?

How are you addressing the environmental impact of AI workloads? Are you using cloud-based tools to measure or reduce that footprint?

Do you believe global regulations are keeping pace with AI innovation, or are companies left to navigate the gray areas on their own? Share your insights in the comments below.

Stay updated with the latest tech news by subscribing to our Tech Today newsletter, delivered to your inbox every morning.

Follow my day-to-day project updates on social media. Subscribe to my weekly update newsletter and connect with me on Twitter/X at @DavidGewirtz, on Facebook at Facebook.com/DavidGewirtz, on Instagram at Instagram.com/DavidGewirtz, on Bluesky at @DavidGewirtz.com, and on YouTube at YouTube.com/DavidGewirtzTV.