CXOToday has engaged in an exclusive interview with Vishal Gupta, Founder, and CEO, Seclore

1. How has Generative AI’s growing popularity impacted the evolution of enterprise offerings, and what are some key use cases where it has been integrated successfully?

Generative AI’s surging popularity has reshaped enterprise offerings in India across multiple dimensions. It promises increased productivity, potentially unlocking USD 621 billion in productive capacity, as estimated by McKinsey. However, it also raises concerns about job displacement, impacting approximately 1% of India’s IT workforce.

Despite challenges, generative AI has found successful integration in various use cases:

1. Content Creation: Generative AI aids in producing news articles, product descriptions, and social media content efficiently.
2. Chatbots: AI-powered chatbots enhance customer service, automate tasks like appointment scheduling, and offer round-the-clock support.
3. Image and Video Generation: Generative AI contributes to creating visuals for advertising and entertainment, boosting creativity.

In India, the rapid adoption of generative AI presents immense potential for economic growth. To harness its benefits, policymakers must prioritize workforce preparation and establish supportive AI policies, fostering innovation and development in the country.

2. Could you elaborate on the data security risks that arise from the expanded attack surface created by the widespread use of Generative AI in various industries? What measures are being taken to mitigate these risks?

The widespread use of generative AI in various industries has expanded the attack surface, leading to several data security risks. These risks include the potential for data overflow, where sensitive business data is stored in third-party spaces, IP leaks that could result in deepfakes or fake information generation, and concerns about data privacy and compliance violations. Additionally, data poisoning, model manipulation, and data theft are emerging threats that must be addressed.

To mitigate these risks effectively, organizations can consider several measures. These include the use of encryption for data in transit and at rest, stringent access controls to limit data access to authorized personnel, proper anonymization of synthetic data, and the implementation of a comprehensive data strategy. Compliance with data privacy regulations is crucial, and securing generative AI models with robust access controls and monitoring is essential to prevent attacks.

In conclusion, while generative AI offers tremendous potential, organizations must remain vigilant about the associated security challenges. Implementing these measures can help safeguard sensitive data and mitigate the risks associated with the expanded attack surface created by generative AI.

3. Considering the increasing reliance on Generative AI and concerns about data security, what role should government regulations play in ensuring both autonomy and data protection in this evolving landscape?

The increasing reliance on generative AI and concerns about data security have raised the need for government regulations to ensure autonomy and data protection. The Indian government’s stance on AI regulation has been mixed, but it must consider several factors:

1. DPDP Act: While the Digital Personal Data Protection (DPDP) Act addresses personal data processing, it lacks explicit AI regulation, potentially conflicting with AI’s power.
2. Digital India Act: The government intends to regulate AI to protect users, likely through the Digital India Act, while sector-specific laws should align with AI deployment.
3. Secure Data Practices: Prioritizing secure data practices and ethical frameworks is vital for data privacy and security.
4. Made in India Stays in India: Emphasizing indigenous AI development aligns with India’s economic and political goals.

In conclusion the government regulations should balance generative AI’s potential benefits with data security risks, considering sector-specific contexts, secure practices, and the “Made in India Stays in India” approach.

4. What are the key reasons for prioritizing the protection of data itself over the infrastructure when working with smart AI systems? How does this shift in focus impact overall security strategies?

Prioritizing the protection of data itself over the infrastructure is paramount in India’s adoption of smart AI systems. This shift in focus is instrumental in upholding data security and addressing potential risks. It impacts security strategies by necessitating a consideration of technical limitations in AI systems, emphasizing data privacy compliance, trust-building, and ethical adherence, and recognizing the critical role of cybersecurity, especially within the AI context. By incorporating these elements into AI system development and operation, organizations in India can promote transparency, data security, and ethical use. The implementation of regulations is essential to prevent privacy erosion, safeguard civil liberties, and mitigate biases, ensuring that AI technology benefits society while preserving individual rights.

5. Can you provide examples of recent incidents or vulnerabilities related to Generative AI that highlight the urgency of addressing data security in this context, and what lessons can be learned from these cases to improve future practices?

While there are no recent incidents or vulnerabilities related to generative AI in India, the importance of addressing data security in this context cannot be overstated. Lessons from discussions on generative AI highlight the need for adaptive security measures that leverage AI’s continuous learning to stay ahead of evolving threats. Deception and honeypots, utilizing generative AI, can divert attackers and gather valuable threat intelligence. Automated incident response through AI-driven SOAR systems improves response efficiency. Awareness of indirect prompt injection attacks on AI systems underscores the necessity of rigorous testing and protection against manipulation. In sum, proactive security practices are imperative to safeguard data in the evolving landscape of generative AI.