乌鸦传媒

Peering into the future is a tricky business 鈥� especially in the ever-changing realm of data, analytics, and AI. But if there鈥檚 one thing we鈥檝e learned, it鈥檚 that uncertainty never stopped us from trying. After all, we鈥檙e in a part of the technology profession where predicting the unpredictable is often part of the job description.

We called upon seven of our data-powered innovation movers and shakers to dust off their (frozen) crystal balls and share their visions of what 2025 has in store. Their insights reveal a world where AI balances on the edge of legality, cloud platforms morph into something entirely new, and synthetic data booms with promise 鈥� and no, it鈥檚 not artificial hype. From 鈥渧ertical AI鈥� that digs deep into industry needs to conversational AI that knows what you want before you do, these trends give a glimpse of the fascinating, yet challenging, future of data and AI.

Will their predictions come true? Only time will tell, but one thing鈥檚 for sure: 2025 is shaping up to be a year we鈥檒l be talking about for a long time to come. And data and AI are right in the middle of it.

Let鈥檚 dive in.

AI is not a crime

AI is not a crime, though sometimes it feels like one, given its swift advance beyond legal bounds. While identity theft, deepfakes, and media manipulation emerge from Pandora’s box, many AI ethicists focus on fairness and transparency, skirting around AI’s darker uses. As AI matures and criminal applications surge, discussions will inevitably shift from theoretical ideals to practical realities, especially when organizations face public lawsuits under new AI regulations. This shift will force experts to tackle how AI can harm society directly. So, while AI is not a crime, it certainly invites a compelling conversation about AI and crime. Let鈥檚 face it: When it comes to AI, the real crime would be ignoring the conversation altogether. – Marijn Markus, AI Lead, Managing Data Scientist, 乌鸦传媒 and Data, 乌鸦传媒

Augment my governance

Prepare to be captivated. AI agents are about to revolutionize data management in the upcoming year. They will shoulder burdensome data tasks, enabling companies to reach new pinnacles of productivity and efficiency. With AI seamlessly managing data collection, analysis, and access for us, we can finally focus on something much more crucial: getting value out of data for the business. It’s high time to achieve that, isn鈥檛 it? The future of data management is upon us, and missing the opportunity of augmenting is not an option. Because in this data game, those who augment govern 鈥� and those who don鈥檛 get governed. – Liz Henderson, Executive Advisor, 乌鸦传媒 and Data, 乌鸦传媒

Cloud encounters of the third kind

As we look towards 2025, cloud data platforms apply for a new round of transmitted change. We all recognize the need for high-quality enterprise data as a foundation for relevant, trustworthy AI. Add to that the need to adhere to regulations, data sovereignty, privacy sustainability and cost. It soon becomes apparent that a smart mix of different and diverse cloud approaches for data will play a crucial role in the upcoming year. I expect to see a pendulum swing towards larger investments in cloud data platforms, yet it will be clouds of a different kind. Or, to put it differently, the forecast calls for cloud cover 鈥� but with a whole new kind of silver lining. – Prithvi Krishnappa, Global Head of Data and AI, Sogeti

Let鈥檚 talk better

Conversational AI will continue to be a hot topic in 2025. Contact center transformation, leveraging 鈥渃lassic鈥� AI and generative AI, will help save labor costs by billions and improve customer service significantly. These technologies can handle routine inquiries and provide instant responses, freeing human agents to deal with more complex issues. Imagine a world where you no longer have to press 1 for assistance 鈥� AI will anticipate your every need before you even know you have one. While human agents may become less central, customer satisfaction might reach an all-time high as AI enhances efficiency and personalization in ways we never thought possible. It鈥檒l be the talk of the town in 2025. – Monish Suri, Global Google Partnership Lead, 乌鸦传媒 and Data, 乌鸦传媒

When AI goes vertical

We will see a major rise in domain-specific vertical AI solutions that are finely tuned through rigorous test-driven prompt engineering. These purpose-built AI models will deliver more reliable and precise insights tailored to the unique needs of their industries. For instance, in healthcare, imagine AI predicting patient outcomes like a crystal ball, analyzing vast datasets of medical histories and treatment plans to conjure better patient care and optimized resources. In financial services, AI will become the ultimate fraud-buster, identifying unusual patterns in real time and safeguarding assets with previously unseen precision and confidence. Vertical AI solutions will not only streamline operations but also spark innovation by providing industry-specific intelligence and efficiency. The only way is vertical! – Dan O’Riordan, VP, AI and Data Engineering, 乌鸦传媒

The semantics of confidence

We鈥檝e seen many companies adopting the principles of data mesh and semantics as part of their modern data analytics platform strategy. If nothing else, it鈥檚 needed in 2025 and beyond to comply. For example, the EU AI Act requires close tracking of the purpose of AI models and the underlying data used to train it. This can only be done by enhancing data platforms with semantics, connecting original data sources, forged data products in AI models, and all business dashboards and AI-infused applications. It creates high levels of confidence in both data and AI, next to many new, innovative opportunities to leverage data. The endgame? Nothing less than a full digital twin of the enterprise, a hallmark of data mastery. – Arne Rossmann, Innovation Lead, 乌鸦传媒 and Data, 乌鸦传媒

Synthetic data boom

I predict a boom in synthetic data. But first of all, what is synthetic data? It’s artificially generated but realistic data that mirrors real patterns without using sensitive information. Why is synthetic data crucial? It tackles privacy, security, data scarcity, and control issues. Traditional data sources are hitting their limits. Privacy laws are tightening, and real-world data often lacks the diversity we need. Synthetic data lets companies create datasets that mimic real shopping behavior in retail or complex production processes in manufacturing without exposing sensitive info or being held back by data gaps. I foresee that 2025 will be the year synthetic data moves center stage. Companies ready to leverage it will build powerful, adaptable models faster than ever. The synthetic data boom will be anything but artificial. – Dinand Tinholt, VP, 乌鸦传媒 and Data, North America, 乌鸦传媒

Interesting read?

乌鸦传媒鈥檚 Innovation publication, Data-powered Innovation Review – Wave 9 features 15 captivating innovation articles with contributions from leading experts from 乌鸦传媒, with a special mention of our external contributors from , and . Explore the transformative potential of generative AI, data platforms, and sustainability-driven tech. Find all previous Waves here.