As AI adoption accelerates, key questions are emerging: How do companies move from experimentation to real impact? What kind of infrastructure is needed to support AI at scale? And how can we ensure this technology is used responsibly, transparently, and in ways that benefit more than just the tech giants?
In this conversation with David Fisher, Industry Principal for Media and Entertainment, EMEA at data cloud company Snowflake, Fisher tells us how media businesses are unlocking the value of their data through AI, and more.
For those of who might not know, what does Snowflake do?
Snowflake is an end-to-end cloud data platform powered by AI, which is easy for customers to use. Customers can store, process, and do analytics on all data types. The platform does data analytics, data collaboration, data engineering, and sharing. Cortex, our AI tooling, is our AI service within the cloud.
What trends are you seeing among users when it comes to deploying AI in production?
We have about 11,000 customers, and 4,000 of them are using machine learning or AI on a weekly basis. They can deploy their AI models on their enterprise data within a secure, governed environment.
We also recently announced OpenAI as a model that our customers can use. We’re the only vendor that offers OpenAI and Anthropic at the moment. So, we have 4,000 customers experimenting in the space. Some of these customers are using it to build chatbots to talk to their data using LLMs. They’re building their own models on the LLMs, doing things like prediction optimisation and ad activation, and they’re developing data from unstructured data insights.
We’ve also made an investment in Twelve Labs, which allows us to deploy multimodal AI on the platform. These multimodal AI models are capable of processing multiple types of data, and can understand videos like a human.
How is AI transforming the creative and operational workflows of agencies and Martech teams?
A lot of Martech is built on, or connected to, us. Users are “mining”, using the models and customer data to do planning, optimisation, cohort building, and activation. Martech is really leaning into AI lately.
The other area is how agencies are using AI. Agencies are really thinking about the areas that are resource intensive and repetitive. Things like media planning and media activation are being done by AI. We’re seeing a convergence of capabilities between independent, mid-sized agencies and the larger holding companies – AI is helping level the playing field.
One of the most exciting developments we’ve seen is the use of AI on customer data sets for dynamic cohort building and activation. AI can surface patterns and insights that would take media planners weeks – or might never be discovered manually. It automatically identifies and names cohorts, generates personas, and activates those audiences into media channels. It’s taking on the heavy lifting of repetitive tasks and accelerating time-to-insight in a way that’s genuinely transformative.
What are the biggest misconceptions businesses have about deploying AI?
As we move through the current AI hype cycle, many organisations are asking important questions – not just about if they should adopt AI, but how to do it in a way that’s enterprise-grade and truly impactful. There’s a growing demand to integrate large language models within secure, scalable software environments that deliver real business value.
We’ve been working closely with the Digital Production Partnership (DPP), which conducts an annual CTO survey. This year, the top priority cited by CTOs wasn’t generative AI for creative, but rather AI for workflow automation. The focus is clearly shifting toward practical, operational use cases.
While AI may eventually transform creative work as well, its most immediate value lies in automating and optimising workflows. Tasks like media planning, campaign optimisation, and insight generation – especially those that are repetitive and time-consuming – are where AI excels. For agencies and marketers, this means faster, more efficient execution and the ability to scale intelligence across the business.
Do you believe AI can finally crack the code on mass personalisation at scale – and what role does data strategy play in that?
Marketing today is an arms race. Whether it’s about selling products or winning share of voice, data is the critical fuel powering it all. The more high-quality data you can feed into AI models, the greater the potential for marketing efficiency and effectiveness.
Personalisation sits at the centre of this opportunity, but it’s also a long-standing challenge. There’s been a pursuit of the ‘holy grail’ – mass personalisation at scale – while still maintaining the reach and impact of broad-based campaigns. Whether AI can finally bridge that gap remains to be seen, but it’s certainly getting us closer.
What we’re starting to witness is the convergence of two data worlds: one from hyper-targeted, personalised advertising, and another from broader broadcast campaigns that traditionally come with higher levels of media waste. When AI can unify insights across both, it opens up entirely new possibilities for marketers to balance precision with scale.
What are the limits of AI when it comes to creativity in TV advertising?
Humans absolutely need to stay in the loop – AI shouldn’t be making final decisions about what goes to broadcast. That responsibility still relies heavily on the expertise and judgement of the agency teams.
In the television ecosystem, for example, there are strict clearance processes that simply can’t be automated. Every ad must go through approval – often app by app – before it can air, making mass personalisation at scale incredibly complex in that context.
That said, AI still has a significant role to play. From audience targeting to creative variation and campaign delivery, it can drive meaningful efficiency and performance. My hope is that we see a smart blend: AI used for what it does best – optimisation and scale – combined with human oversight and creative strategy. We’re committed to supporting large media organisations, whether they operate in broadcast, digital, or print, as they explore AI-powered solutions for advertising and marketing sales.
With AI reshaping the media value chain, who do you think is best positioned?
I think everyone can benefit. I think the risk is that the largest technology companies, who are also advertising businesses, take the lion’s share of the benefits. That’s why most advertising spending goes into the largest tech platforms. So, they will benefit because they have the investments to make the most of AI.
In terms of agencies, media owners, and publishers, I think who will benefit is who has the most data. So, I think marketers will benefit because they’re sitting on enormous amounts of first party data. Agencies can benefit because they are operating as a third party between the marketer and the advertising platforms, they can benefit from the collaboration and deployment of AI into their client’s data, and can understand what happens in advertising environments. My hope is that media owners – television businesses and content businesses – will find the benefits of AI as well for that point of media plurality. AI is something that can be used by all enterprises.
What is your biggest concern surrounding AI?
My biggest concern with AI isn’t job loss. In fact, I believe AI has the potential to enhance productivity and create entirely new kinds of roles, as we’ve seen with every major technological shift. The real challenge lies elsewhere. One of the most pressing issues is sustainability, both in terms of energy usage and the long-term impact of scaling these technologies. But perhaps even more concerning is the growing reliance on AI systems that operate as black boxes, especially within the major hyper-scalers. We’ve already seen this dynamic unfold with social media algorithms, where opaque systems shape public discourse without transparency or accountability.
Now, we risk repeating that pattern in advertising. If hyperscalers control the AI behind ad targeting, optimisation, and measurement – without broader understanding or oversight – it could create an uneven playing field. That’s why it’s so important for those of us in media and advertising to help level the field: by enabling more organisations, not just the tech giants, to build and deploy advanced AI models of their own. That’s how we ensure competition, innovation, and fairness in this next era.
