News related to
intelligence

High-Quality Growth with Improved Gross Profit Margin, increasing R&D Investment and Strategic Focus on Trustworthy Agents Ecosystem

HONG KONG SAR – Media OutReach Newswire – 27 March 2026 – Voicecomm Technology Co., Ltd. (“Voicecomm Technology” or the “Company“, Stock Code: 2495.HK), a core technology provider and ecosystem operator of trustworthy conversational AI in China, announced its annual results for the year ended December 31, 2025. Leveraging its deep technological expertise and forward-looking strategic initiatives in trustworthy conversational AI, the Company achieved steady revenue growth, continuous improvement in profitability, and a significant enhancement in operating cash flow, further strengthening its competitive edge in the enterprise-level trustworthy Agent sector.

Gross Profit Margin Steadily Improves, Operating Cash Flow Turns Strongly Positive

In 2025, the Company’s total revenue successfully exceeded the RMB1 billion mark, reaching RMB1,006.9 million, representing a year-on-year increase of 7.0%. Gross profit amounted to RMB551.2 million, increased by 8.0% from the previous year, while the gross profit margin increased by 0.5 percentage point to 54.7% from 54.2% in the same period last year, reflecting the high-value-added products and technical services of the Company, as well as effective cost control.

Notably, the Company’s net cash generated from operating activities turned strongly positive, reaching a net inflow of RMB212.5 million, compared to a net outflow of RMB129.2 million in the same period last year. This improvement underscores enhanced operational efficiency and reflects strengthened receivables management, which accelerated cash collection.

Profit for the year amounted to RMB140.2 million. The significant increase in net profit compared to the same period last year was mainly attributable to eliminating the impacts of changes in carrying amount of redeemable capital contributions, an accounting adjustment arising from financing agreements entered into with shareholders prior to the Listing and completion of the Global Offering.

Increasing R&D Investment to Strengthen the Trustworthy Agent Technology Foundations

As a technology-centric product company, Voicecomm Technology remains committed to the independent R&D and innovation of underlying technologies. In 2025, the Company’s research and development expenses reached RMB224.3 million, representing a substantial year-on-year increase of 67.7%. The investment was primarily directed towards developing Agents with continuous learning capabilities and a technical framework for Multi-agent collaboration, aiming to enhance technological capabilities and elevate the level of product innovation. This reinforces our trustworthy Agent technical architecture composed of three layers: “Multimodal Perception + Multi-model Thinking + Multi-agent Collaboration”.

This architecture takes the “meta-model” as the core to effectively alleviate the common pain points of in enterprise-level implementation, such as hallucinations, controllability and data security, by integrating the generalization capabilities of large language models with the precision of vertical domain knowledge. On this basis, the Company have formed a deliverable and operable trustworthy Agent product system, to ensure that the Agents are usable, manageable, and controllable in enterprise environments, thereby powering the intelligent upgrade of six core application scenarios: City management and administration, Automotive and transportation, Telecommunications, Finance, Healthcare, and Energy management.

Productization Strategy Drives Deep Application across Six Core Scenarios

Voicecomm Technology focuses on empowering various industries through superior products and technologies. In 2025, leveraging its mature product matrix, the Company successfully established replicable benchmark standards across multiple application scenarios.

City Management and Administration: As a leading solution provider in the smart government sector in China, the Company’s business has covered more than 130 prefecture-level cities. The “Smart Government Agent” deeply integrates the capacities of large language models, enabling more intelligent and automated government services with standardized and intelligent applications in scenarios such as government hotlines and city governance.

Automotive and Transportation: The Company successfully established benchmark autonomous driving projects in cities such as Mianyang, Zibo, and Ezhou, building a successful and replicable “Smart Transportation Agent” solution. For the Mianyang Science and Technology New City project, a total of 96 autonomous vehicles have been deployed, and the project was successfully selected as a National AI Application Pilot Testing Base in the transportation sector. In January 2026, the Company newly won the bid for the “Ezhou Huahu Airport Smart Port” autonomous driving bus procurement and operation project, further expanding its application boundaries.

Healthcare: In January 2026, the Company successfully won the bid for the “Chuannan Smart Valley AI Vertical Large Model Innovation Platform – Silver Economy Construction and Operation Project” in Neijiang City, Sichuan Province, with a total contract value of nearly RMB300 million. This project represents the Company’s first “AI + Elderly Care” city-level benchmark demonstration project. It adopts a closed-loop collaborative model of “online platform + offline service network + home terminals,” integrating Agents capabilities with the Company’s “vertical small model microservices” system in areas such as Health Early Warning, Cognitive Ability Assessment, and emotional companionship into a productized solution, thereby establishing rapidly replicable city-level smart elderly care operational benchmark.

Telecommunications, Finance, and Energy Management: The Company continues to deepen collaborations with leading enterprises in the telecommunications and finance sectors, leveraging the “Telecommunication Service Agent” and “Financial Service Agent” to enhance service efficiency and user service value. Meanwhile, its AI-powered smart charging solution has been progressively integrating charging pile networks in China and across several Southeast Asian countries.

Benefiting from the successful implementation of the Company’s productization strategy and the high level of market recognition for its trustworthy Agent solutions, as of December 31, 2025, the Group’s project pipeline and orders in hand saw significant growth. The number of ongoing projects at year-end increased to 320, representing a year-on-year increase of 41.6%, while the outstanding contract sum at year-end rose to RMB1,048.9 million, a year-on-year increase of 57.4%, reflecting the continued expansion of the Company’s business.

Future Outlook: Focusing on the Trustworthy Agents Ecosystem with Four Strategic Priorities

Looking ahead, the Company will firmly focus on its goal of “building a trustworthy conversational AI ecosystem” and will advance the following strategic priorities:

Overall and Technology Strategy: Continue to focus on R&D and innovation in frontier technologies such as multi-modal fusion and trustworthy intelligence, promoting the deployment of trustworthy agents across more application scenarios. By creating open technology platforms and standards, the Company will attract more developers and partners to jointly build a prosperous and win-win industrial ecosystem.

Market Strategy: Establish benchmarks for quality and innovation within the industry and deepening partnerships with various service channels. At the same time, the Company will actively expand into the C-end market, extending cutting-edge technologies to a wider user base, thereby expanding the influence and commercial value of the ecosystem.

Regional Strategy: Domestically, the Company will continue to deepen its partnerships with major cooperating cities to create smart city benchmark cases. Internationally, the Company will actively respond to the “Belt and Road” initiative, grasp the tremendous potential of emerging markets, and promote the Company’s trustworthy Agent products and services globally to enhance its international brand image.

Investment Strategy: Through prudent strategic investments and M&A, the Company will optimize the layout of the upstream and downstream industry chains and consolidate the stability and competitiveness of the ecosystem.

DR.Tang Jinghua, Chairman ofVoicecomm Technology Co., Ltd., said: ” 2025 was a landmark year for Voicecomm Technology. We not only achieved a strong turnaround in operating cash flow and a steady increase in gross profit margin financially, but we also completed a strategic leap at the technological and business level towards becoming a ‘a core technology provider and ecosystem operator of trustworthy conversational AI.’ We deeply understand that the essence of enterprise-grade AI lies in creating replicable and reliable products and technology foundations. During the year, we significantly increased R&D investment and successfully applied our trustworthy Agents across six core scenarios, particularly in city-level benchmark projects in emerging fields like smart elderly care, showing the strong competitiveness of our productization strategy. Looking to the future, we will continue to pursue the goal of ‘building a trustworthy conversational AI ecosystem,’ leveraging an open platform to gather ecosystem partners and empowering diverse industries with innovative technology, thereby creating long-term sustainable value for our shareholders and society. ”

Hashtag: #Voicecomm

The issuer is solely responsible for the content of this announcement.

Voicecomm Technology Co., Ltd.

Founded in 2005 and headquartered in Wuhan, Voicecomm Technology is one of the leading core technology providers and ecosystem operators of trustworthy conversational AI listed on the Main Board of the Hong Kong Stock Exchange, and obtained the qualification as National-level SRDI, Technology Little Giant and High-Tech Enterprise. Leveraging its proprietary trustworthy Agent，the Company overcomes pain points in the commercialization of large language models, such as hallucinations and compliance issues, ensuring AI is usable, manageable, and controllable in enterprise environments. For enterprise-level customers, the Company provides AI services covering the entire process of “communication –decision – execution”. Its trustworthy Agent solutions have been widely deployed across multiple scenarios, including City management and administration, Automotive and transportation, Telecommunications, Finance, Healthcare, and Energy management, and it is dedicated to empowering the intelligent transformation of diverse industries.

Introducing new security solution designed for the rapidly emerging era of agentic AI

HONG KONG SAR – Media OutReach Newswire – 25 March 2026 – TrendAI™, the enterprise AI security leader, today introduced TrendAI™ Agentic Governance Gateway, a new security solution engineered to give organizations visibility and control over autonomous agent interactions to strengthen security where systems interact across data, tools, and environments with increasing autonomy.

To learn more about TrendAI™ and TrendAI™ Agentic Governance Gateway, visit: trendaisecurity.com.

Rachel Jin, CPBO and Head of TrendAI™: “Tools like OpenClaw show just how powerful and accessible this new model has become. Organizations need to deploy these systems to unlock the next wave of productivity, and many already are, often without centralized oversight. TrendAI™ Agentic Governance Gateway enables this by providing visibility, control, and confidence.”

Traditional cybersecurity is built to protect endpoints, networks, and applications. In contrast, agentic AI systems operate across dynamic chains of interaction where agents, models, APIs, and data continuously exchange information and trigger actions.

Eva Chen, CEO of Trend Micro: “As AI systems become more autonomous, security must evolve from protection to governance. This is the next frontier of cybersecurity and the focus of TrendAI™.”

TrendAI™ Agentic Governance Gateway is a new way to address the emerging security gap created by agentic AI systems, such as OpenClaw, where autonomous agents act across enterprise environments without clear security control points. This was demonstrated last week at NVIDIA GTC, where a fundamentally new attack surface was highlighted – one that traditional security models were not designed to control: Autonomous agentic frameworks, such as OpenClaw, accelerating enterprise adoption of AI systems capable of planning, executing, and coordinating actions across workflows at machine-speed. There has never been a more critical business-level need than understanding and governing how agentic systems behave, and what actions they are taking. The orchestration required to monitor this attack surface created complex, multi-step workflows across enterprise systems until now.

According to Forrester, “AI agents are proliferating across workflows, but security programs built for human-centric architectures fail in agentic environments. These agents operate with dynamic reasoning, ephemeral identities, and goal-driven autonomy, creating unpredictable attack paths. Risks to agentic architectures include intent hijacking and cascading hallucinations that extend beyond confidentiality to integrity and availability. Without guardrails, enterprises risk regulatory violations, financial loss, and disclosure events from agentic security issues.” ¹

TrendAI™’s Agentic Governance Gateway allows enterprises to focus on the behavior, interactions, and outcomes of AI systems in real-world environments. TrendAI™ Agentic Governance Gateway is delivered through the TrendAI Vision One™ platform, building on TrendAI™’s existing strengths, including AI-driven analytics to detect anomalous behavior and emerging threats, and a unified approach that correlates context across endpoints, cloud, applications, and AI systems.

TrendAI™’s Agentic Governance Gateway enables enterprises to:

• Gain visibility into how agents interact across systems
• Understand the context and intent behind agent communications to identify risky or unintended actions
• Enforce policy and control over agent-driven actions
• Introduce human oversight at critical decision points
• Simulate governance decisions before deployment — previewing the full policy impact without executing it
• Stage, preview, and roll back governance changes through a managed lifecycle

With these new advanced capabilities, TrendAI™ is enabling organizations to secure the critical interaction layer – the dynamic communication fabric where autonomous systems coordinate, make decisions, and drive enterprise actions. By establishing robust oversight and control at this pivotal layer, TrendAI™ ensures that agentic interactions are visible, governed, and trusted, effectively closing the security gap in agent-driven AI environments.

TrendAI™ is the enterprise business unit of Trend Micro Incorporated (TYO: 4704; TSE: 4704).

1: The AEGIS Agent-On-A-Page Template For Agentic Security, Forrester Research, Inc., Feb 13, 2026.
Hashtag: #trendai #trendmicro #trendvisionone #trendaivisionone #visionone #cybersecurity

https://www.trendaisecurity.com
https://www.linkedin.com/company/trendai-security
https://x.com/trendaisecurity
https://www.facebook.com/trendaisecurity/
https://www.instagram.com/trendaisecurity/

The issuer is solely responsible for the content of this announcement.

About TrendAI™

TrendAI™, a global leader in AI security, empowers enterprises to innovate fearlessly by securing AI, cloud, networks, endpoints, and data across the modern attack surface. At the core is TrendAI Vision One™, a unified cybersecurity platform that centralizes cyber risk exposure management and security operations to protect the entire AI lifecycle from infrastructure to models to users. The platform is fueled by world-class threat intelligence and insights that protect organizations from hundreds of millions of threats every day. With 6,000 TrendAI™ experts across 75 countries, TrendAI™ empowers security leaders to stay ahead of threats, driving proactive security outcomes across the entire attack surface. This includes critical environments like AWS, Google, and Microsoft. AI Fearlessly.

The purchase, attributed to cargoes loaded through intermediaries and structured channels, underscores a shifting pattern in global oil flows as refiners seek cost advantages amid volatile benchmarks. Market participants said the acquisition reflects both commercial pragmatism and a broader recalibration of sourcing strategies by large refiners navigating supply disruptions and pricing differentials.

Reliance, which operates one of the world’s largest refining complexes at Jamnagar, has historically processed Iranian crude due to its favourable refining characteristics and pricing structure. The latest move suggests the company is again exploring opportunities linked to Iranian grades, which are often discounted relative to comparable Middle Eastern benchmarks.

Iran’s oil exports have been subject to US sanctions since Washington withdrew from the nuclear agreement in 2018. Those restrictions significantly curtailed Tehran’s ability to sell crude openly, pushing exports into opaque channels involving ship-to-ship transfers, reflagging, and complex payment mechanisms. Despite these constraints, Iran has continued to move significant volumes to Asia, particularly to buyers willing to navigate compliance risks.

Industry analysts say the renewed interest in Iranian oil stems from tightening margins and evolving global supply dynamics. Production curbs by OPEC+ members, combined with geopolitical disruptions affecting Russian and Middle Eastern supplies, have altered trade routes and pricing structures. Refiners are increasingly prioritising flexibility, often sourcing from multiple regions to optimise margins.

Reliance’s procurement strategy has expanded notably over the past two years, with the company becoming a major buyer of discounted crude, including shipments linked to Russia following Western sanctions. The latest Iranian cargoes appear to align with this broader approach, where cost competitiveness outweighs traditional sourcing patterns.

Energy traders indicate that the cargoes in question were likely acquired through intermediaries rather than direct state-to-company contracts, a common practice in sanctioned markets. Such arrangements can involve layered ownership structures, making it difficult to trace the origin of the oil definitively. Payment mechanisms also tend to bypass conventional banking channels, relying on alternative currencies or barter-like systems.

Officials in Tehran have repeatedly stated that oil exports remain a central pillar of the country’s economic resilience, even under sanctions. Output has gradually recovered, with production estimates rising above 3 million barrels per day in recent months. Much of this increase has been driven by exports to Asia, where demand remains robust despite regulatory scrutiny.

Global oil prices have remained sensitive to geopolitical developments, with Brent crude fluctuating amid concerns over supply security and demand recovery in major economies. The availability of discounted barrels from sanctioned producers has created a parallel market that exerts downward pressure on certain price segments, benefiting refiners capable of accessing such supplies.

For Reliance, the ability to process a wide range of crude grades provides a strategic advantage. Its refining configuration allows it to handle heavier and more complex crudes, including those from Iran, Venezuela, and Russia. This flexibility enables the company to capitalise on arbitrage opportunities that may not be available to less sophisticated refiners.

Policy implications remain a key consideration. While there has been no formal indication of a shift in sanctions policy by Western governments, enforcement patterns have shown variability. Analysts note that enforcement intensity often fluctuates depending on broader diplomatic priorities, creating windows of opportunity for market participants.

New Delhi has maintained a delicate balance in its energy diplomacy, emphasising the need for affordable supplies while adhering to international frameworks. Purchases of crude from sanctioned producers have typically been structured to avoid direct violations, with companies relying on legal and compliance teams to manage exposure.

Shipping data and market intelligence suggest that Iranian crude continues to reach global markets through a network of vessels operating under flags of convenience. These ships often disable tracking systems or conduct transfers in international waters, complicating monitoring efforts.