AI leaders warn of cyber risk & human oversight gaps

Technology and security leaders are using AI Appreciation Day to highlight tightening cyber risk, data quality concerns, and the growing need for human oversight of artificial intelligence in business.

The comments reflect rising unease about the speed of AI-driven cyber threats and the maturity of corporate adoption, even as investment in the technology accelerates.

Anthropic's recent Mythos disclosure findings have sharpened industry focus on how quickly attackers can weaponise newly disclosed software vulnerabilities with AI tools. The findings suggest exploit development windows have shrunk dramatically, now falling well within the response cycles many organisations still rely on.

James Greenwood, AVP Solutions Engineering APAC at Tanium, said rapid AI-enabled exploitation is undermining traditional patching and incident response practices built around longer lead times after a Common Vulnerabilities and Exposures disclosure.

"Anthropic's Mythos disclosure findings have made it clear: the window from CVE disclosure to weaponisation is now measured in hours, not days, and certainly not the weeks businesses have planned around. AI is already building working exploits from known vulnerabilities in hours for under US$50. Most businesses and CISOs simply aren't ready for what's to come. Technical capability is not the obstacle. Organisations need to get out of their own way, shifting from a posture of manual reaction to one of autonomous control, with their people focused on the decisions that actually require human judgement. Organisations need to move to an autonomous remediation state where routine patching, compliance validation, and vulnerability closure happen at machine speed, with humans setting policy, reviewing exceptions, and approving high-risk changes. That shift is only possible when security teams have complete, real-time visibility and control of every endpoint across the estate," said James Greenwood, AVP Solutions Engineering APAC at Tanium.

Security and risk teams are under pressure to automate more defensive workflows as attackers use generative AI to assemble exploits and phishing campaigns at low cost. Greenwood's comments underscore a broader industry shift toward real-time endpoint awareness and machine-driven remediation.

Alongside security concerns, business leaders are examining how AI is reshaping work and decision-making. Some executives emphasise that the technology still depends heavily on human framing and governance.

Ruhee Meghani, Chief Executive Officer and Founder of Allied Collective, said AI's role as an amplifier of human choices raises questions about leadership, trust, and workplace impact.

"AI is rapidly transforming the way we work, learn, solve problems, and relate with other humans. Promising higher returns, increasing productivity, and analysing data in seconds, it's got the world in a chokehold, with businesses rushing to adopt AI. Sure, it might seem like the silver bullet to usher us into the next new age of technology. However, AI shouldn't be celebrated in isolation. AI is an amplifier of human input. Behind every AI tool are human decisions about how it's designed, trained, and used. That means its impact isn't determined by the technology alone, but by the choices we make in how we use it. With the jury still out on whether it moves the needle on productivity gains and wellbeing, the question is whether AI will enhance human capability or replace it. It can't build trust, navigate difficult conversations with curiosity, or understand the cultural and social context behind a decision using critical thinking. That's why organisations need thoughtful leadership and strong governance to ensure AI supports better, not just faster, decision-making. As AI increasingly becomes part of everyday business, uniquely human qualities become more valuable. The best leaders won't just ask, 'should we use AI?' They'll also ask, 'How does this impact our future?' and 'Who might this affect?' The future won't be defined by technology alone, but by how responsibly, inclusively, and thoughtfully we choose to use it," said Ruhee Meghani, Chief Executive Officer and Founder of Allied Collective.

Data foundations and information management are also coming under scrutiny as organisations expand AI pilots into production-scale projects. Industry observers argue that many firms underestimate the work required to prepare data across both physical and digital formats.

Garry Valenzisi, Vice President and General Manager at Iron Mountain ANZ, said Australian organisations in particular show a gap between enthusiasm for AI and the data practices needed for reliable outcomes.

"AI is transforming the way organisations work, helping unlock new levels of productivity, insight, and innovation. But while everyone wants the promise of AI, far fewer are confronting the data reality required to deliver it.

"Australia is widely regarded as being among the world's leaders in AI adoption, yet it is often considered to lag in AI sophistication and maturity. This highlights a critical challenge across Australian industries: organisations are racing to adopt AI while attempting to scale it on fragmented information, disconnected systems, and inconsistent governance.

"When the information feeding AI cannot be trusted, neither can its outputs. Meanwhile, critical information that could fuel growth remains trapped in legacy systems. As organisations move from AI experimentation to enterprise deployment, these information challenges are becoming barriers to delivering meaningful business value.

"On AI Appreciation Day, it's worth recognising that AI's greatest potential doesn't come from increasingly powerful models alone. It depends on information that is accurate, well-governed, and available to the people and systems that need it. When organisations can confidently connect, govern, protect, and activate information across physical and digital environments, AI has a far stronger foundation for delivering meaningful business outcomes. The organisations that succeed in the AI era will be those that invest in these foundations today, building the operational confidence needed to scale AI and translate ambition into measurable business outcomes," said Garry Valenzisi, Vice President and General Manager at Iron Mountain ANZ.