As AI becomes more and more integrated into everyday life, industry leaders and experts are predicting a transformational 2025.
From groundbreaking developments to existential challenges, the evolution of AI continues to shape industries, alter workflows and spark deeper conversations about their meaning.
In this article, we have confirmed that AI News has caught up with some of the world’s major minds and imagines it for the next year.
Smaller, purpose-driven model
Grant Shipley, senior director of AI at Red Hat, predicts a transition from the assessment of AI models with substantial parameter counts.
“2025 will be the year when the model will stop the number of numbers of a certain parameter as a metric indicating the value of the model,” he said.
Instead, AI focuses on specific applications. Developers migrate to chain small models in a way similar to microservices in software development. This modular, task-based approach may facilitate more efficient, bespoke applications that suit your specific needs.
The open source that leads the way
Bill Higgins, VP of Watsonx Platform Engineering and IBM’s open innovation, expects the open source AI model to grow in popularity in 2025.
“Despite rising pressure, many companies struggle to show measurable returns on AI investments. The high licensing fees in proprietary models are a major factor. In 2025, open source AI solutions have been It manifests as a dominant force in filling this gap,” he explains.
In addition to the affordable price of open source AI models, it also provides greater transparency and customization possibilities, making it ideal for multi-cloud environments. Open source models can provide a way for businesses to move beyond experiments and to scalability, as they match their own systems to their power.
This is a prediction from Nick Burling, Nasuni’s SVP who believes 2025 will lead to a more measured approach to AI investment.
“Companies will focus on strategic use of AI, ensuring that all AI initiatives are justified by clear and measurable returns,” Burling said.
Cost efficiency and edge data management become important, helping organizations optimize their operations while keeping their budgets low.
Enhance human expertise
For Turing CEO Jonathan Siddharth, the standout feature of the 2025 AI Systems is the ability to learn from large-scale human expertise.
“The key advancements come from teaching AI not only what to do, but how to approach problems with logical reasoning that coding naturally cultivates,” he says.
Competitiveness, especially in industries such as finance and healthcare, depends on integrating this integration of human expertise with AI.
Behavioral psychology catches up
Understanding the interaction between human behavior and AI systems is at the forefront of predictions by Niklas Mortensen, Chief Design Officer at Designit.
“There are so many examples of algorithmic bias that lead to unnecessary production, and behavioural psychology will catch up with AI trains because humans are human,” explained Mortensen.
Solution? An experiment of “pause moments” for human surveillance and intentional balance between automation and human control in critical operations such as healthcare and transportation.
Mortensen also believes that individual AI assistants will ultimately prove their worth by fulfilling their long-standing possibilities to organize our lives efficiently and intuitively.
The bridge between the physical and digital worlds
Andy Wilson, senior director of Dropbox, envisions that AI will become an integral part of our daily lives.
“AI will evolve from being a useful tool to becoming an integral part of everyday life and work. It offers innovative ways to connect, create and collaborate,” says Wilson.
Mobile devices and wearables are at the forefront of this transformation, offering a seamless, AI-driven experience.
However, Wilson warns of new questions about the boundaries between personal and workplace data that are spurred by such integrations.
Promote sustainability goals
With businesses’ sustainability goals for 2030, with IBM’s VP ESG & Asset Management Kendra Dekeyrel, we highlight how AI can bridge the gap.
Dekeyrel is urging organizations to adopt AI-powered technologies to manage energy consumption, lifecycle performance, and data center tensions.
“These features will ultimately help advance our overall sustainability goals,” she explains.
Unlock calculus and inference
James Ingram, Streetbees’ VP technology, foresees the changing computational requirements as AI handles increasingly complex problems.
“The focus shifts to inference calculations before training,” he said, highlighting the importance of real-time inference capabilities.
Extending the context window also greatly enhances the way AI holds and processes information, potentially outweighing human efficiency in a given domain.
The rise of the fundamentals of agent AI and unified data
According to Dominique Wellington, enterprise architect at Snaplogic, “Agent AI marks a more flexible and creative era for AI in 2025.”
However, such systems require robust data integration, as siloed information risks its reliability.
Wellington expects 2025 to witness advanced solutions to improve data hygiene, integrity and systematics. Everything is essential for Agent AI to thrive.
From hype to reality
Jason Shane, Cognit’s field CTO, predicts that 2025 will be remembered as the year in which truly transformative and validated generator AI solutions emerge.
“The singular example of embedding truly transforms Gen AI into a real workflow through the fog of AI for AI stands out,” predicts Schern.
These domain-specific AI agents revolutionize industrial workflows by providing customized decision-making. Sheln cited an example in which AI reduced its time-consuming root cause analysis from months to just minutes.
Deepfake and the crisis of trust
Sophisticated AI threatens the reliability of images, videos and information, and Gen.
“Even experts may not be able to tell you the real thing,” warns Stephen Nisson.
To combat this crisis, robust digital credentials are required to verify trustworthiness and promote trust in increasingly blurred digital reality.
2025: Basic changes in AI landscape
As multiple predictions converge, it is clear that fundamental changes are on the horizon.
Experts who contributed to forecasting the industry this year highlighted smarter applications, stronger integration with human expertise, closer alignment with sustainability goals, and enhanced security. However, many also foresee important ethical challenges.
2025 represents an important year. The transition from the initial excitement of AI proliferation to mature and measured adoption that promises a value and a more nuanced understanding of its impact.
See: AI Action Summit: Leaders call for unity and fair development
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo in Amsterdam, California and London. The comprehensive event will be held in collaboration with other major events, including the Intelligent Automation Conference, Blockx, Digital Transformation Week, and Cyber Security & Cloud Expo.
Check out other upcoming Enterprise Technology events and webinars with TechForge here.
