AI Laws are Coming: Are Indian Enterprises Ready for Responsible Automation? | TechTree.com

AI Laws are Coming: Are Indian Enterprises Ready for Responsible Automation?

Scaling with Scrutiny: Why the Next Phase of AI is Built on Oversight

 

By Darshil Shah

There was a time when adopting AI felt like a clear advantage. Teams moved quickly, pilots turned into live use cases, and systems began to influence decisions in ways that were not always visible from the outside. That momentum has not slowed. Reports continue to show that 47% of Indian enterprises are already running multiple AI use cases, with many planning to increase their investments further.

But now compliance requirements are mirroring this growth. The Digital Personal Data Protection Act has brought sharper focus on how data is collected and used. At the same time, the government has begun outlining expectations around AI governance, especially in areas where automated systems shape outcomes or generate content. The shift is subtle, but it is steady. Enterprises are no longer only being asked how fast they can scale AI. They are being asked how well they understand and explain what their systems are doing.

When Adoption Moves Faster Than Oversight

Across industries, AI has moved beyond experimentation. In many organisations, it now sits inside daily workflows. Around 15% of organisations have already deployed AI extensively across business units and customer touchpoints, and nearly 48% rely on a mix of SaaS, OEM, and custom-built models. This level of integration did not happen overnight, but it did happen quickly enough that oversight structures often struggled to keep pace.

In several cases, governance was treated as something to be addressed later, once systems proved their value. That approach worked when AI was limited to smaller use cases. It becomes harder to sustain when those systems start interacting with each other and influencing larger parts of the business. Even now, close to 60% of organisations have only basic governance policies, and just 19% have carried out structured risk assessments across legal, ethical, and societal aspects. These issues have begun to garner attention as questions around accountability start coming up more often. This brings back focus on how data moves across the system and different departments within the company.

Control Struggling to Keep Up With Scale

The absence of unified oversight does not end with how AI systems are deployed. It becomes more visible in how prepared enterprises are to manage what follows. As use cases expand across systems that were never designed to work together, readiness often begins to lag behind adoption. What appears as scale on the surface does not always reflect control underneath.

Recent findings bring this gap into focus. Around 71% of organisations struggle to interpret DPDP requirements, while 39% already see compliance as a key barrier. At the same time, nearly 94% plan to increase AI spending, even as only a small share have conducted structured risk assessments. This contrast suggests that readiness is still catching up, and that managing AI responsibly now depends less on deployment and more on how well enterprises can account for it.

When Outputs Start Drawing Attention

The question of enterprise readiness does not end with how prepared organisations are to manage data flows. Things get more complicated when attention shifts to what AI systems actually produce and how that output is handled. Recent changes under the Information Technology Intermediary Guidelines and Digital Media Ethics Code Rules 2021 and its 2026 amendments place responsibility on enterprises to identify synthetic media, enable labelling, and respond to misleading outputs when needed. This adds pressure on systems that are still finding their footing in terms of control and visibility.

For enterprises, it is not just about how well a system performs. It also comes down to showing how an output came together and whether it meets basic safeguards. This becomes more sensitive in customer-facing interactions, where automated responses directly influence trust. As these expectations build, isolated controls start to fall short, and governance begins to feel more layered and difficult to manage in a connected environment.

Building Structure Into Everyday Operations

What begins as a compliance requirement often turns into an operational question. It becomes less about meeting a rule and more about how systems stay aligned when data, decisions, and outputs are all moving at once. Many enterprises are finding that this cannot be addressed through separate tools or periodic checks. It needs to be built into the way operations are managed.

This is where platforms like ERP start to play a different role. Not as a standalone system, but as a layer that brings consistency across processes. When information flows through a connected structure, it becomes easier to see how it is being used. Data trails are clearer. Decisions are easier to follow. Processes do not have to be reworked each time a new requirement comes in. Over time, this reduces the effort needed to stay aligned with evolving expectations, while still allowing systems to scale in a more controlled way.

Conclusion

The shift underway may not appear dramatic on the surface, yet its implications are becoming harder to set aside. AI will continue to expand across functions, but the conditions around its use are becoming more defined through evolving governance frameworks and enforcement signals. This is shaping how organisations think about scale, not just in terms of speed, but in terms of how well systems can hold up under scrutiny.

Enterprises that respond early are beginning to build with these expectations in mind, aligning their operational choices with emerging regulatory direction rather than reacting to it later. This approach does not limit progress. It allows it to move with greater clarity. Over time, the ability to operate with that clarity will influence how steadily organisations can navigate a space where both innovation and accountability are becoming equally consequential

 

(The article is written by Darshil Shah, Founder and Director, TreadBinary, a TechCon, and the views expressed in this article are his own)

 


TAGS: Artificial Intelligence, dpdp act

 
News Corner
IMP IMP IMP
##