Security Threats, AI Reliability Top 2025 Software Development Challenges
There is no shortage of challenges that organizations need to deal with regarding modern software development.
That’s the general theme of the sixth annual Reveal 2025 Top Software Development Challenges survey from Infragistics. The report identifies digital trust as the paramount concern for tech leaders in 2025, marking a significant shift from previous years’ challenges around AI integration and resource limitations. It found that organizations are now grappling with the consequences of rapid digital transformation, including heightened security vulnerabilities, questions about AI reliability, and an intensifying talent shortage for specialized roles.
Key findings of the survey include:
-
Security (51%), AI code reliability (45%), and data privacy (41%) rank as the top software development challenges for 2025.
-
73% of tech leaders plan to expand AI use within their organizations in the coming year.
-
78% of tech leaders expect to increase their focus on business intelligence in 2025.
“One of the biggest challenges identified by this year’s Reveal survey was AI code reliability, with 45% of respondents citing it as one of their top concerns,” JJ McGuigan, Reveal and App Builder Product Marketing Manager, Infragistics, told ITPro Today.
The Challenge of AI for Software Development
Adding further insult to injury, despite widespread AI adoption, 55% of tech leaders view AI deployment as their biggest business challenge.
McGuigan noted that organizations face unanticipated challenges in AI deployment, even as its adoption grows. One major barrier is the difficulty of integrating AI systems with existing infrastructure, which often requires significant adjustments to legacy technology. Another challenge is the lack of quality data, as many AI models rely on large, clean datasets that organizations struggle to obtain or maintain.
“Our survey reveals that many CIOs and CTOs underestimated the complexities involved in AI deployment, discovering that simply adopting the technology isn’t enough,” McGuigan said.
The key to success is to have a plan. McGuigan said that early adopters who rushed to integrate AI without a cohesive plan are now dealing with fragmented systems, operational inefficiencies, and inconsistent performance.
According to McGuigan, successful AI implementations tend to prioritize clear objectives, invest in continuous team training, and ensure robust data governance practices. In contrast, struggling implementations often lack these foundations, resulting in missed opportunities and inefficiencies.
Why Digital Trust Is a Top Challenge
The survey highlights digital trust as the critical concern, with security (51%) leading the challenges for 2025.
Security threats in modern software development have evolved significantly, with increasingly sophisticated attacks targeting vulnerabilities in code, cloud infrastructure, and third-party dependencies.
For example, McGuigan said supply chain attacks, like the SolarWinds breach, which triggered a much larger supply chain incident that affected thousands of organizations, including the U.S. government, highlight the risk of malicious code.
In response, organizations are adopting proactive measures that go beyond traditional cybersecurity approaches. Those approaches include embedding security within the development life cycle through DevSecOps practices.
“They are also utilizing automated threat detection tools powered by AI to identify and mitigate potential risks in real time, enhancing their ability to swiftly respond to emerging threats,” McGuigan said.
Skills Gap Continues to Grow
A key issue impacting all forms of IT, not just software development, is the ability to hire staff with the right skills.
Among the reasons why cybersecurity is increasingly difficult is the growing shortage of skilled professionals in the cybersecurity field. According to the Reveal report, cybersecurity engineer ranks as the third most difficult technical role to fill, underscoring the severity of this talent gap.
AI engineers are now the most sought-after technical professionals, with 28% of tech leaders reporting difficulties filling these positions — nearly double the demand for the next most needed roles: IT security engineers (16%) and cybersecurity engineers (13%). This represents a significant shift from 2024, when software developers (25.4%), data analysts (14.6%), and web designers (11.7%) topped the hard-to-fill positions list.
“Unfortunately, in many cases, AI is advancing more rapidly than organizations’ ability to train and reskill their employees on this evolving technology,” McGuigan said. “Companies that want to succeed need to adapt to the growing demand for AI engineers, invest in AI training initiatives and reskilling programs, and expand their recruitment efforts.”
link
