playnext.fm

CDO Magazine Podcast Series · CDO Magazine

PODCAST | A CTO’s View on Why Enterprise AI Still Breaks at the Data Layer

February 12, 2026·26 min·2 clips
CTO Jyoti Chawla reveals why AI projects break at the data layer and what truly gives companies an edge.
AI keeps moving fast. Jyoti argues that data is the steadier place to build advantage because models, vendors, and application frameworks change too quickly to carry the whole strategy. Karen starts with the practical question: how should organizations think about data if they want AI value that lasts? Jyoti's answer has layers. Semantic understanding and runtime inspection help teams ask whether an AI decision still matches the business goal. Human review is not treated as a nice blanket safeguard. It is part of the machinery that checks meaning and intent. Security also gets a harder job. Once AI systems reason over data and make decisions, the threat model has to include semantic validation, not only access and storage. Controls still matter, but Jyoti talks about them in working categories: detective, preventative, and mitigative controls across the AI lifecycle. Regulation can help set guardrails, but it cannot carry the design. Use cases and data minimization do more of the daily work. Privacy has to start early, with privacy by design, differential privacy, encryption, and firm limits on data use. Encryption expands from data at rest and in motion to confidential computing for data in use. Auditing matters because AI actions need a trail when companies trust them inside enterprise workflows. Jyoti also wants clearer labels for models, much like nutritional facts for data. By the end, trust is the adoption story: enterprise platforms, trusted partners, rapid change, and agentic AI that is still moving fast.

As heard by us

Enterprise AI still depends more on the data layer than the model layer.

CDO Magazine treats enterprise AI as a data and control problem first, and Jyoti Chawla keeps that point in view. Her case is plain enough: vendors, models, and frameworks change fast, so the lasting work sits in data quality, semantic understanding, and human oversight rather…

Read the full review in PlayNext →

Why you'd press play

You want a CTO's practical read on why data, controls, and oversight matter in AI.

Read the full recommendation in PlayNext →
Listen to the show on
Apple PodcastsSpotify
Play clips from this episodeMore from CDO Magazine Podcast Series