2025 Tech Trends Maturity
and Suitability Index Survey

Large-scale AI, which trains on large datasets, has become one of the hottest topics in security. Especially, it is featured in various Chinese video security solutions which are claimed to provide better detection, lower false alarm rates and smarter video search. Yet in our survey, the technology only received moderate scores of 3.98 and 3.49, respectively for suitability and maturity. Indeed, being new in security, large-scale AI still sees several technical barriers that need to be overcome.

“As our experience suggest, users often perceive these technologies as not yet fully mature. One reason is the need for extensive real-world data to train robust, generalizable AI models, which is a challenge in highly variable, high-security environments. Additionally, especially in mission-critical scenarios, organizations are cautious about relying exclusively on large-scale AI until consistency and transparency are further enhanced,” said Hanchul Kim, CEO of Suprema.

Yet in the long run, the technology is expected to see increased adoption due to its benefits for users.

“Large-scale AI represents a powerful evolution in security where AI systems can learn from massive, diverse datasets across sites, use cases, and conditions. In theory, this allows algorithms to become far more context-aware and adaptive than traditional, site-specific models that are disconnected and in silos. However, in practice, we’re still in an early phase,” said Philippe Henaine, Manager of Strategic Partners at i-PRO.

He adds: “The technology itself is promising and available, but real-world integration into monitoring workflows remains complex due to factors like latency, reliability, and cybersecurity. Today, vision-language models (VLM) already transform how operators interact with video data, turning complex analytics into proactive insights. Large-scale AI models will follow, enabling cross-site intelligence and predictive capabilities once infrastructure and data governance fully catch up.”

On the contrary, AI analytics for business intelligence scored pretty high in the survey, receiving 4.33 and 3.87 for suitability and maturity. AI analytics for business intelligence are advanced analytics that can extract actionable insights from video and other types of data. Examples include object detection and tracking, facial recognition, behavior analysis and heat mapping. With such analytics, users can optimize staff allocation, understand customer behavior and detect risks before incidents occur – all of which can help improve the brand image, customer loyalty and sustainability of the business. Given these benefits, we can expect strong demand for AI analytics for business intelligence down the road.

Cybersecurity scored the top 4.45 and 4.13, respectively, for suitability and maturity. Indeed, at a time when networked devices are increasingly connected and when video can be easily faked or manipulated, cybersecurity will only gain more relevance and urgency.

“The risks surrounding cybersecurity today are very real. With the rise of generative AI tools, video footage can be manipulated, stolen, or shared without authorization, putting both organizations and individuals at risk. Protecting video data and being able to prove its integrity has become critical, especially as we generate more data and insights,” Henaine said. “Today, all our camera models already meet FIPS 140-2 Level 3 standards, ensuring a high security baseline. Building on this, we are moving toward standardizing the highest level of cybersecurity, FIPS 140-3 Level 3, across our entire lineup by the end of next year. From entry-level to flagship models, every product will feature secure boot and firmware, and advanced encryption for both data and video streams.”

“Cybersecurity remains a top priority as threats evolve alongside technology. Customers value security solutions that offer strong, standardized data protection, regulatory compliance, and rapid response to emerging attacks,” Kim said.

“Suprema has proactively obtained certifications like CE RED and ISO/IEC 27001/27701, incorporating secure coding, robust encryption, access controls, and continuous monitoring across cloud, network, and device layers,” he continues. “Our architectural approach adheres rigorously to privacy-by-design principles, complying with GDPR regulations, the EU AI Act, and NDAA requirements. Investing in advanced cybersecurity is essential not just for regulatory reasons, but to uphold user trust and ensure uninterrupted, resilient operations in an increasingly digital security landscape.”

IoT or the Internet of Things also scored highly in the survey, gaining 4.31 and 4.11 in suitability and maturity. Needless to say, today IoT and security are so intertwined it’s nearly impossible to separate them. It’s because of IoT that security was able to branch off into smart home, smart building, smart transportation, smart retail, smart city and other “smart” applications. When IoT is combined with AI, the resulting AIoT technology plays a critical role in helping users secure premises, gain business intelligence, and boost operational/management efficiency. We can say the impact of IoT on security has been significant, and their influence on each other will only continue.

5G was another top scorer, garnering 4.21 and 3.99 for suitability and technical maturity. This is understandable as 5G increasingly plays a key role in security. 5G allows high-speed data transmission and covers longer distances than Wi-Fi. This makes it suitable for remote areas or moving objects where laying cables is difficult. In fact, more and more 5G cameras are now available in the market. With 5G becoming more mature and widespread, it will gain more traction in security over time.