AI will be central to building a more accountable and trustworthy media measurement ecosystem: Tata Elxsi

As media consumption fragments across television, OTT, connected TVs and digital platforms, audience measurement is becoming increasingly complex. In this interview, Nitish Kumar, Global Practice Head – Technology, Tata Elxsi, discusses how artificial intelligence is transforming media measurement by enabling cross platform audience insights while ensuring transparency, 
explain ability and trust.

He also explains why governance, human oversight and responsible AI will be critical to the future of advertising and audience analytics.

AI is increasingly being used to measure audiences, campaign effectiveness and consumer engagement. How is it changing the media measurement landscape today?

AI is changing media measurement by shifting the focus from simply reporting audience numbers to understanding audience behaviour across an increasingly fragmented media landscape. Consumers today move between linear television, OTT platforms, connected TVs, mobile devices, websites and social media throughout the day. Hence, measuring each platform on its own no longer gives broadcasters and advertisers a complete picture of how audiences consume content or engage with brands.

AI brings the ability to process and correlate large volumes of data from these different touchpoints, helping broadcasters, advertisers and content owners build a more comprehensive view of audience behaviour. Rather than looking only at what was watched, content owners can now see how audiences engage across platforms, how viewing patterns change over time and where campaigns are delivering maximum impact. This matters even more as new viewing behaviours continue to emerge. Co-viewing during live sporting events, out-of-home viewing in restaurants or public venues and multidevice consumption are all becoming common and much of this falls outside traditional household-based measurement models. AI helps close these gaps by bringing together fragmented audience signals into insights that broadcasters and advertisers can actually act on. 

As the industry moves towards cross-platform measurement, AI will have an important role to play in enabling this broader audience intelligence, provided transparency, consistency and 
trust remain central to how the measurement process is built.

Today, data is coming from different sources, including TV, OTT platforms, and digital channels. How can broadcasters and media companies ensure that this data remains accurate, transparent and trustworthy throughout the entire measurement process?

The industry today has no shortage of data. The real challenge is ensuring that data collected from multiple platforms is consistent, accurate and trustworthy. Audience signals now come from television panels, connected TVs, OTT platforms, mobile applications, websites, social media and return-path data and each source follows its own collection methodology and measurement standards. Without strong governance, inconsistencies can quickly spread across the measurement chain and affect business decisions. 

Building trust starts with full visibility into the data lifecycle. Media companies should be able to trace where data originates, understand how it was collected, know what transformations were applied and explain how AI models contributed to the final outcome. Every audience metric should be explainable, auditable and backed by clear governance.

AI has an important role to play here too, monitoring data quality, flagging anomalies, spotting unusual viewing patterns and checking measurement consistency across platforms. As the 
industry moves towards unified cross-media measurement, AI can also help reconcile audiences across TV, OTT and digital channels while reducing duplication, giving advertisers a clearer basis for investment decisions.

The ultimate measure of success is building a transparent and reliable view of audience behaviour that broadcasters, advertisers and agencies can confidently rely on.

Many traditional media companies are still cautious about using AI for audience measurement and analytics. What is the first step they should take to build confidence in AI-driven systems?

The first step isn't a large-scale AI transformation programme. It is identifying a focused business problem where AI can demonstrate measurable value. Audience measurement has always been a high-trust discipline because it directly influences advertising investments, content strategy and business performance. Any new technology entering this ecosystem has to earn credibility through proven outcomes.

A practical approach is to start with use cases such as audience segmentation, campaign effective ness analysis, data quality monitoring or anomaly detection. These areas allow organisations to evaluate what AI can do without disrupting existing measurement frameworks.

To build stronger confidence in AI, running it alongside current methodologies allows organisations to compare outcomes, check performance and understand where AI actually delivers
improvement.

Transparency matters just as much. Business users should know what data has been analysed, how conclusions were reached and what limitations exist in the models. When stakeholders are able to explain the results rather than simply accepting them, confidence grows. 

Practically, AI adoption should be viewed as an enhancement rather than a replacement for established systems. As organisations see consistent improvements in measurement quality and operational efficiency, trust builds naturally, creating the foundation for wider adoption across the media value chain.

Looking ahead, do you believe AI systems will eventually be able to audit themselves, or will human oversight always remain essential for ensuring trust and accountability?

AI systems will keep becoming more capable of monitoring and validating their own performance; however, the process will continue to require human oversight. We're already seeing real progress in areas such as automated model governance, anomaly detection, model drift monitoring and ongoing quality checks, and these capabilities will only get stronger as the systems mature.

In media measurement, AI will increasingly spot inconsistencies across datasets, validate measurement assumptions, compare outputs against historical viewing patterns and flag unusual  results for further investigation. This will improve operational efficiency as well as the overall quality and reliability of measurement.

That said, accountability can't be handed over entirely to technology. Audience measurement directly influences advertising investment, content strategy and commercial outcomes across multiple stakeholders. Human oversight remains essential when interpreting audience behaviour, resolving conflicting data sources, understanding regional or cultural viewing patterns and making final business decisions.

So, the future isn't likely to be fully autonomous. It will instead be a collaborative model, with AI acting as a monitoring and validation layer, continuously checking data quality and flagging risks, while human expertise continues to provide governance, accountability and final decision-making.

How can media measurement companies balance the need for innovation with increasing expectations around privacy, compliance and ethical AI use?

Innovation and responsibility shouldn't be viewed as competing priorities. Sustainable innovation is only possible when it's built on real trust. As media measurement relies more on  behavioural signals collected across multiple devices and platforms, organisations need to show that privacy, transparency and accountability are built into their systems from the start.

This starts with privacy-by design: privacy must be built in from the outset rather than added afterwards. Data collection needs to be transparent; consent has to be communicated in plain language and personal information should be anonymised where required. Governance must also keep pace with changing regulations, backed by regular audits and clear lines of accountability.

Responsible innovation should not come at the expense of compliance. Organisations also need to ensure that what AI is telling them reflects real audience behaviour and isn't influenced by bias or synthetic data that could skew decisions. As measurement stretches across TV, OTT and digital, trust itself starts to matter as much as the numbers; it becomes something companies compete on.

The organisations getting this right – pairing innovation with ethical AI, sound governance and careful data handling – are the ones advertisers, broadcasters, regulators and audiences  will trust most.

India is increasingly emphasizing responsible AI through frameworks focused on transparency and accountability. How do you see these principles influencing the future of media measurement and analytics? 

India's growing focus on responsible AI comes at an important time, as AI becomes a bigger part of audience measurement, advertising analytics and consumer engagement. As organisations 
rely more on AI-generated insights, transparency and accountability stop being optional and become essential.

One important piece of this is data lineage. Organisations need to clearly show how data moves from its source through every stage of processing before it becomes a final measurement 
output. This kind of transparency lets stakeholders understand how audience metrics are actually generated, which in turn builds confidence in the underlying methodology. 

The industry is also likely to see growing demand for explainable AI and audit-ready measurement  platforms. Advertisers and media companies will increasingly expect visibility into how metrics such as reach, engagement and audience behaviour are calculated, especially as digital consumption continues to evolve beyond traditional television measurement.

India's highly diverse media landscape presents a real opportunity to build responsible AI frameworks that account for complex, multi-platform viewing behaviour while remaining fair, accountable and transparent. Organisations that invest early in explainable methodologies and strong governance  will be well placed to build trust and establish long-term credibility in an increasingly AI-driven media ecosystem.

How are you helping media and entertainment companies navigate the growing demand for trustworthy, transparent and explainable AI solutions? 

At Tata Elxsi, we believe successful AI adoption requires equal focus on business outcomes, engineering excellence and trust. Our experience across digital video platforms, OTT ecosystems, connected devices, content moderation and data engineering enables us to help media organisations build AI solutions that deliver measurable value while maintaining trust in the underlying data.

Our work spans audience intelligence, content operations, metadata enrichment, consumer segmentation, churn prediction, campaign analytics and decision support. Beyond expanding AI capabilities, we also help organisations establish governance frameworks that align with regulatory requirements and support greater transparency throughout the AI lifecycle. 

One of the industry's biggest challenges today is bringing together fragmented data across multiple technology platforms. Billing systems, content platforms, consumption data and customer insights often operate in silos, which limits an organisation's ability to build a unified view of the business. AI and modern data platforms help orchestrate these disconnected systems, creating a more consistent and trusted picture of audience behaviour.

Our goal isn't simply to deploy AI, but to help media organisations innovate with confidence – making sure transparency, explain ability and governance remain central to every solution we build.

From Tata Elxsi's perspective, what opportunities do you see for AI to create greater confidence and accountability across the advertising and media value chain?

The greatest opportunity for AI lies not just in automation and operational efficiency, but in strengthening trust across the entire advertising and media ecosystem. Today's audiences consume content across television, OTT platforms, connected TVs, mobile devices and social media, making it challenging to build a consistent picture of audience behaviour through traditional measurement approaches.

AI can bring together fragmented data from these different environments while continuously validating data quality, spotting anomalies, detecting fraudulent activity, reducing duplication and improving measurement consistency.

Looking ahead, one of the industry's long-standing aspirations is a trusted, cross-platform view of audiences. No single measurement methodology can fully capture today's complex viewing behaviour, but AI offers a way to bridge these gaps by bringing multiple data sources into a more unified and transparent framework.
This gives advertisers, broadcasters and media owners a more reliable and comprehensive foundation for audience intelligence and decision-making. 

Looking ahead, one of the industry's long-standing aspirations is a trusted, cross-platform view of audiences. No single measurement methodology can fully capture today's complex viewing behaviour, but AI offers a way to bridge these gaps by bringing multiple data sources into a more unified and transparent framework.

We at Tata Elxsi strongly believe that AI will play a central role in building a more accountable, explainable and trustworthy media ecosystem – one where every stakeholder, from  broadcasters and advertisers to agencies and regulators, can have greater confidence in the insights driving their most important business decisions.

Author: Nitish Kumar, Global Practice Head – Technology, Tata Elxsi