Resilient data governance and risk control are key to accelerating safe AI at scale

SANDEEP BHAMBURE
VP & MD (INDIA & SAARC), VEEAM SOFTWARE

“AI is in the air, everyone is just talking AI. While nearly 50% of large enterprises already have AI use cases live, what really lacks is corporate governance. That’s why I’m talking about Accelerating Safe AI at Scale. The intent is huge — India Inc. is going to spend $200 billion over the next three years - but there is definitely an execution risk. If organizations can control governance, security, and resilience, AI projects can really be successful.

In the AI era, whatever technologies and solutions were adopted to implement resilience no longer work. Organizations need to look at resilience and security with a completely different lens. The top risks are data privacy, legal compliance, governance capabilities, and the quality, consistency, and observability of data. Most of these risks are data-centric. Observability is critical — you need to know what kind of data is where in your organization, whether it is sensitive, who can access it, how it will be analyzed, and whether you are breaching regulatory compliance like DPDP. As long as these are taken care of, you can have confidence in driving your AI projects. India has the potential to be not just a consumer of AI but also someone who can build its own AI initiatives. Talent is abundant, young professionals are engaging, and many businesses are being started by them. But here comes the AI paradox — the friction between innovation and risk. There is pressure to accelerate AI, but the speed of innovation gets slowed down by the risks associated with AI products.

Safe AI at scale is built on three pillars. First, a resilient data foundation — immutable backups, zero-trust architecture, and multi-cloud protection cannot be afterthoughts. Second, intelligent protection — anomaly detection, behavioural threat monitoring, and AI itself helping uncover threats proactively. Third, responsible and compliant AI — real-time governance, continuous audits, and recoverability of AI workloads. You need to understand your data, secure it for AI, build resilience in your AI infrastructure, and activate your AI agents responsibly. Preparation for DPDP compliance aligns closely with feeding data into AI models. By integrating data resilience, governance, and compliance, enterprises can accelerate AI initiatives while managing risk. If your data is resilient, you can confidently accelerate your AI initiative, knowing your AI is in safe hands.”