VMblog: AppSOC Security Predictions for 2025

* Read the VMblog article *

Industry executives and experts share their predictions for 2025.  Read them in this 17th annual VMblog.com series exclusive.

By Willy Leichter, CMO, AppSOC

The new year begins with several notable issues and trends in cybersecurity:

• Ransomware continues to dominate: This is both predictable and frustrating. Given that this has been a continuous security focus for years, you would expect more progress. The problem is that ransomware is not a specific technique - it's more of a proven way to monetize most types of breaches. Extortion has always been effective and there's little indication that this will change.
  

• AI goes mainstream, along with security concerns: In the last year, AI usage and projects went from being mostly experimental to being widely deployed and adopted. While it is still early days, security experts are beginning to reckon with a much larger attack surface area, moving from just code vulnerabilities (which are still alive and well), to MLOps, LLMs and other models, datasets, notebooks, and more.
  

• Software supply chains remain a weak link: Over the last few years, there have been an increasing number of high-profile supply chain attacks - such as SolarWinds and Log4j. While there is more awareness of this problem, we're not close to solving it because the dependence on third-party code continues to expand. Addressing supply chain security requires considerably better communication and cooperation with suppliers, and much greater scrutiny of third-party code than most security teams can handle.
  

Here are some predictions for security issues in 2025:

• AI offense will have an edge over AI defense: We know that AI will be used increasingly on both sides of the cyber war. However, attackers will continue to be less constrained because they worry less about AI accuracy, ethics, or unintended consequences. Techniques such as highly personalized phishing and scouring networks for legacy weaknesses will benefit from AI. While AI has huge potential defensively, there are more constraints - both legal and practical, that will slow adoption.
  

• AI systems will become targets: AI technology greatly expands the attack surface area with rapidly emerging threats to models, datasets, and MLOps systems. Also, when AI applications are rushed from the lab to production, the full security impact won't be understood until the inevitable breaches occur.
  

• Security teams will have to take charge over AI security: This sounds obvious, but in many organizations, initial AI projects have been driven by data scientist and business specialists, who often bypass conventional application security processes. Security teams will fight a losing battle if they try to block or slow down AI initiatives, but they will have to bring rogue AI projects under the security and compliance umbrella.
  

• Supply chain exposure will expand: We've already seen supply chains become a major vector for attack, as complex software stacks rely heavily on third-party and open-source code. The explosion of AI adoption makes this target larger with new complex vectors of attack on datasets and models. Understanding the lineage of models and maintaining integrity of changing datasets is a complex problem, and currently there is no viable way for an AI model to "unlearn" poisonous data.
  

• Lessons learned - AI brings its own risks: 2024 saw the beginnings of the broad realization of AI's promise and potential, evidenced both by major breakthroughs and increasing experimenting with use cases, accompanied in many cases by the launch of public-facing applications. It was also the year when many realized that AI and LLM systems introduce new risks, and - equally problematic -- that security teams are often only vaguely aware of fast-moving AI projects.
  

‍

The Road Ahead

Blocking important AI projects will not work and will put companies at a competitive disadvantage. Instead, organizations must enable AI innovation by ensuring adequate visibility, guardrails, application security, and governance to prevent costly and damaging security incidents.

How can organizations address these threats and begin to more fully realize the innovation and growth potential that AI puts within reach?

Meeting these threats head-on and adopting AI with confidence will demand trusted security.  With AI's move to the production app mainstream, organizations are increasingly understanding that it's no longer sufficient to merely detect new, isolated anomalies. It's imperative to protect AI systems from end-to-end with a complete AI and Application Security platform. This means finding, implementing and embracing a complete solution to discover, correlate, prioritize, remediate, and govern AI systems. This is why AppSOC's focus is covering the new AI attack surface, while providing robust security management for all AI stacks, applications, infrastructure and innovation.

Published Friday, January 31, 2025 7:33 AM by David MarshallFiled under: VMBlog Info, Contributed, Predictions 2025

‍

‍