Newsletters




10 Predictions for AI Development in 2025


AI innovation in the big data space shows no signs of slowing down, with companies rushing to adopt or create their own solutions based on AI, automating redundant tasks, and spinning up all kinds of use cases to justify spending big bucks in the area.

Here, tech experts share their predictions for AI in 2025:

Enterprises will turn to AI upskilling to survive: There are simply not enough AI-centric developers to fulfill the increasing demand for deploying enterprise-grade AI applications. Highly specialized AI developers also come with incredibly high salaries, limiting participation in the AI talent war to a small group of organizations that have the resources to battle for top talent. As a result, enterprises will increasingly turn to internal upskilling initiatives to prime their existing developers so that they can thrive in the era of AI. Upskilling programs will no longer be a "nice to have" perk for employees—they'll be a "need to have" for maintaining a competitive pace of innovation.—Jeff Hollan, head of applications and developer platform, Snowflake

By the end of 2025, many security operations teams will have an AI analyst on their team thanks to the growing focus on agentic AI across industries: They'll have some kind of AI agent that joins their team to help them cope with the volume of alerts. This agent would oversee all the mundane and repetitive tasks that nobody likes to do, helping humans to extend their capacity work on more strategic things, rather than just chasing false positives. As for how we will get there? Some security teams will build an AI agent internally and some will hire vendors to bring these AI agents onto their teams. Either way, we are going to see AI agents helping SOC teams in a big way in 2025.—Itai Tevet, CEO and co-founder of Intezer

Retaining extensive data sets will become essential: Generative AI depends on a wide range of structured, unstructured, internal, and external data. Its potential relies on a strong data ecosystem that supports training, fine-tuning, and Retrieval-Augmented Generation (RAG). For industry-specific models, organizations must retain large volumes of data over time. As the world changes, relevant data becomes apparent only in hindsight, revealing inefficiencies and opportunities. By retaining historical data and integrating it with real-time insights, businesses can turn AI from an experimental tool into a strategic asset, driving tangible value across the organization.—Lenley Hensarling, technical advisor, Aerospike

MLOps and LLMOps combine into AIOps: Machine learning and prompt engineering are very distinct techniques with different tools and use cases. However, they're all part of the broad desire to build robust automation systems. As practitioners get a better handle on the strengths and weaknesses of both, and how they can be used together, expect to see a new wave of frameworks and platforms for leveraging both technologies as one.—Paul Barba, chief scientist, InMoment

Using AI to reduce costs: We expect companies to keep 2025 budgets flat due to uncertainty around inflation, import tariffs, and unemployment levels due to political change. That means companies will continue to focus on cost saving projects like pushing vendors for lower prices, keeping headcount low, and using AI to reduce cost will become more popular.—Alex Levin, CEO of Regal

The rise of Small Language Models: Small Language Models (SLMs) will take off as a means of solving more targeted problems with greater cost efficiency. We need to be more discriminating in what problems we ask LLMs to solve. Many natural language processing (NLP) applications can be solved using more cost-efficient models such as GPT-4o-mini, Gemini-flash, etc. Using more cost-efficient models means lowering the cost-covering point for the use of LLM services.—Joe Regensburger, VP of research at Immuta

AI will move closer to independence: AI will continue to evolve, expanding its interactions in the virtual world by autonomously controlling computers and in the real world through autonomous vehicles, drones, and other IoT devices. This will increase the level of AI-controlled interactions in our lives and even between machines, which will create a whole new layer of ethical, security, and governance concerns that we have yet to experience.—Miguel Baltazar, VP of developer relations, OutSystems

Quantum computing isn’t ready for prime time: Despite the hype, quantum computing won’t hit the mainstream by 2025. Although significant investments are being made in quantum computing, the technology will remain in early innovation stages through 2025, with commercialization still several years away.—Bruce Kornfeld, chief product officer, StorMagic

AI and the fusion of emerging technologies: AI’s true potential lies in its connections with other emerging technologies. While AI itself is transformative, its impact multiplies when combined with quantum computing, intelligent edge, zero trust security, 6G technologies and digital twins, to name a few. This fusion creates a dynamic environment ripe for innovation and addressing existing challenges.—CTO/CAIO John Roese, Dell Technologies

Weaponized AI will be the biggest security concern in 2025—and IT teams will be hit hardest: The biggest security threat we're seeing is the continual evolution of AI. It’s getting really good at content creation and creating false imagery (i.e. deepfakes) and as AI gets better at data attribution, it will become even more difficult for organizations to distinguish between real and malicious personas. Because of this, AI-based attacks will focus more on targeting individuals in 2025. Most of all, IT teams will be hit hardest, due to the keys they possess and the sensitive information they have access to. Most AI-based attacks will target individuals to solicit access and money, and IT organizations need to ensure they’re prepared, educating staff, and shoring up defenses accordingly. The best way to reign in AI risks is with more employee training. People must know what to be on the lookout for, especially as AI technology evolves. In general, you can’t do enough cyber awareness training. It's very real—even beyond AI, there are a ton of ways to compromise an individual system or information, and I think the more that we can educate people, rather than try to curtail the technology, the better.—Mike Arrowsmith, chief trust officer at NinjaOne


Sponsors