Tech leaders call for tighter AI governance & security

Technology leaders are using AI Appreciation Day to call for tighter governance, clearer business outcomes, and stronger security foundations for artificial intelligence deployments.

The annual awareness day comes as organisations embed AI more deeply into operations and begin confronting rising costs and new forms of risk.

Shane Buckley, President and Chief Executive Officer at Gigamon, said AI now sits at the centre of how enterprises operate, rather than at the edge of experimentation.

"AI Appreciation Day comes at an important moment in the technology's evolution. We've moved beyond asking whether AI belongs in the enterprise. The conversation is now centred on how organisations operate in the Mythos era, where increasingly capable AI models and agents are changing the economics of work, innovation, and cybersecurity.

Capabilities that once required specialised expertise, significant time, or large teams are becoming dramatically more accessible. That creates tremendous opportunity, but it also changes how organisations compete, innovate, and manage risk.

The growing cost of AI is also becoming a barrier to widescale deployment. Given the massive infrastructure investment from leading AI companies, token costs are expected to continue rising, putting even more pressure on IT budgets and forcing important investment trade-offs.

As AI becomes embedded across business-critical systems, leaders need a clear understanding of how models, agents, and AI-powered workloads interact with their data and infrastructure. Visibility into AI activity is becoming just as important as the AI itself. It gives organisations the context they need to strengthen security, govern these systems responsibly, make better decisions, and improve cost and ROI controls so spending aligns with return.

The organisations that create lasting value from AI will be the ones that invest as much in visibility, governance, and operational discipline as they do in the technology itself."

Executives across software and data businesses describe a shift from proving AI adoption to tracking concrete returns, particularly in mature markets such as Australia, New Zealand, and Singapore.

Jarrod Kinchington, Vice President and General Manager, APAC, at Smartsheet, said leaders are scrutinising outcomes rather than usage metrics.

"AI Appreciation Day is a useful moment to step back from the hype and recognise how quickly the conversation has matured. From where I sit, the focus has clearly shifted from whether to adopt AI to how to scale it, govern it, and translate it into measurable performance.

What's emerging now is a much more pragmatic mindset. Organisations are no longer trying to prove they are using AI; they are focused on what it is actually delivering. The questions I hear most often from senior executives are grounded and outcome-focused: what is improving, where is value being created, and how is it changing the way work gets done?

Most businesses are already seeing pockets of individual productivity gains, but faster outputs do not automatically translate into stronger organisational performance. The real gap is organisational adoption. Too many organisations are still layering AI onto existing workflows rather than redesigning processes, decisions, and operating models around it.

Across APJ, this plays out differently by market. Australia and New Zealand show strong executive intent and experimentation, but scaling outcomes remains uneven. Singapore tends to align AI more closely to execution.

Across the region, organisations are increasingly disciplined about linking AI to measurable business impact rather than simply driving usage. That means focusing on outcomes over activity, sharing proven use cases, encouraging experimentation within clear guardrails, and measuring productivity, quality, and customer impact.

Despite these differences, the core challenge is consistent: moving from experimentation to outcomes while managing limited execution capacity and competing priorities.

At the same time, AI is reshaping how teams collaborate. Work is evolving from human-to-human to human-to-human-to-AI, with specialised agents supporting everything from research to execution. In this environment, human judgement becomes more important, particularly in areas like critical thinking, decision-making, and communication.

The organisations making the most progress are treating AI as an operating model transformation, not a software rollout. They are investing in data, governance, and leadership capability and, importantly, using governance to enable scale rather than constrain it.

What stands out to me about this moment is that AI is no longer just accelerating how people work; it is forcing organisations to rethink how work happens and how it translates into real business outcomes."

Risk and security specialists are also focusing on the intersection of AI agents, identities, and data as deployments expand across multiple systems and clouds.

George Harb, Vice President, Australia and New Zealand, at OpenText, said many enterprises remain heavily focused on productivity while underestimating the new security pathways AI introduces.

"As enterprises embrace AI at scale and it becomes more embedded in business processes, resilience will depend on three capabilities: understanding what AI can retrieve, controlling how it interacts with systems, and governing the data it consumes.

Many organisations focus on AI performance, productivity gains, and automation opportunities. Equally important is understanding the new breach pathways that emerge when AI is connected to data, user credentials, and business workflows at scale.

AI agents are becoming active participants that can search, retrieve, analyse, and act across enterprise systems. The challenge is that these agents often inherit user privileges from employees, service accounts, or connected applications. Without strong governance, they can aggregate sensitive information across multiple systems and expose data in ways never anticipated.

To prevent AI agents from retrieving data they shouldn't, organisations should assess whether these agents operate within the same security and oversight framework as human employees. Traditional identity and access management was designed for people. AI introduces a new class of digital actors that continuously request information, perform tasks, and interact with systems.

As more agents are deployed, organisations face increasing volumes of automated requests, delegated privileges, and machine-to-machine interactions. Security frameworks built around periodic reviews and manual approvals can quickly become overwhelmed. This results in permission sprawl, dormant credentials, and security pathways that are difficult to monitor or audit. In many cases, existing security and risk processes struggle to keep pace with the scale and speed of automated access.

Another consideration is that many enterprises are connecting AI solutions to data spread across cloud platforms, SaaS applications, file shares, collaboration tools, and legacy systems. When data management practices are inconsistent, models can be trained on and retrieve information from data stores that lack proper classification, ownership, and security controls. As a result, AI may surface information that should never have been available.

This highlights the importance of strong oversight of the quality, security, and provenance of the data feeding AI systems. The next generation of AI risk lies at the intersection of AI, digital identity, and data governance.

As AI becomes more embedded in core business processes, organisations must ensure the foundations supporting it are equally intelligent, secure, and accountable. As organisations mark AI Appreciation Day, the focus should extend beyond AI's capabilities to the systems, identities, and data that will determine whether it can be deployed securely, responsibly, and at scale."