AI leaders call for stricter governance & oversight

Technology leaders are using AI Appreciation Day to call for stricter governance and clearer human oversight as artificial intelligence moves deeper into business operations. Executives from security, software and commerce firms warn that unmanaged autonomy and weak controls could undermine the value AI is creating.

The comments come as so-called agentic AI systems gain the ability to make decisions, execute workflows and act across multiple applications with limited human intervention. Organisations are under growing pressure from regulators and customers to show how they are managing the risks.

Shiv Agarwal, Co-Founder and Chief Executive Officer at Singulr AI, said many companies still rely on outdated governance approaches that do not match the speed of current deployments.

"AI Appreciation Day is the perfect moment to move past the hype and understand how AI can actually create value inside an enterprise. That value is not simply created through better models, more tools or more agents. AI creates value when people can use it to eliminate painstakingly repetitive tasks, make better decisions, experiment safely and securely, and move faster. The challenge in 2026 is that most organizations are trying to govern a constantly evolving, dynamic technology with processes built for a much slower generation of software and controls that have become fragmented over time. Employees are using these new tools before IT and security teams have had a chance to review them. New AI features appear in SaaS applications without as much as a notification. On top of that, agents are starting to take actions across systems. The cost of using AI is skyrocketing without proper usage oversight and optimization controls. Governance cannot just live in a policy or onboarding document anymore; it has to operate in real time, where AI is actually being used. True AI appreciation means recognizing both the technology's promise and the reality of how it operates within a given organization. To get the most value out of AI, your company needs visibility, clear digital and interpersonal controls, training, and confidence to keep internal policies aligned with the real world as AI continues to gain more autonomy," said Agarwal.

Security specialists argue that the governance gap is widest where AI already touches sensitive data and mission-critical systems. Their concern is that enthusiasm for automation is outpacing investment in accountability.

Jay Bavisi, Founder and Group President at EC-Council, framed the issue as a test of leadership and accountability rather than a technology choice.

"AI Appreciation Day is a useful moment to recognize how deeply AI has entered everyday work. But appreciation should not be confused with uncritical adoption. The more mature conversation is about what it takes to adopt AI responsibly, defend the systems it touches and govern its use with discipline. AI is no longer only a productivity tool that helps people write, search or analyze faster. It is moving into workflows where it can recommend actions, trigger processes, access data and influence decisions. That changes the responsibility for every organization using it. The question is not simply how much AI can improve productivity, but whether leaders understand where it is being used, who is accountable for it, how it is secured and when human judgment must intervene. For AI to be sustainable, organizations have to look beyond immediate efficiency gains. The real measure is whether AI can be scaled without weakening trust, increasing unmanaged risk or leaving people unprepared for the decisions these systems now influence. A system that is powerful but poorly understood, widely used but weakly governed, or difficult to secure will eventually create more pressure than progress. The real test for enterprises is not whether they can deploy more AI. Most already can, and many already are. The test is whether their people are prepared to work with AI, question its outputs, defend against its misuse and govern autonomous action before it creates business risk. That is what AI Appreciation Day should remind us of. The future of AI will not be shaped by enthusiasm alone. It will be shaped by the discipline organizations build around AI: Adopt. Govern. Defend," said Bavisi.

In software development, attention is turning to the integrity of AI-generated code and the provenance of automated changes moving into production environments.

Hari Srinivasan, Vice President of Products and Strategy at Lineaje, said the shift toward agentic AI is changing basic assumptions about enterprise risk.

"Agentic AI is driving enterprise technology past 'just' assistance and directly into autonomous operational execution. As these systems independently generate code, execute workflows and make real-time decisions, they fundamentally alter the enterprise risk profile. The shift demands an immediate evolution in governance. Enterprises can no longer rely on periodic audits; they require continuous, real-time visibility into agentic AI, including AI-generated software, and rigorous integrity verification throughout the development lifecycle. That process begins with software lineage-equipping organizations to trace exactly what AI created, verify its origin and validate its integrity before a single line of code reaches production. For AI Appreciation Day, the industry conversation must mature from AI adoption to strict AI governance. As agentic AI reshapes enterprise operating models, continuous governance should become the baseline, not the exception. True value is not what AI can build; it is ensuring that AI and AI-generated outcomes can be governed, verified and trusted," said Srinivasan.

Security leaders in regulated sectors describe similar pressures as they use AI for monitoring, incident triage and compliance evidence.

Ryan Heidorn, Chief Technology Officer at C3, said defence contractors face a growing gap between regulatory expectations and the workload confronting security teams.

"On AI Appreciation Day, I think less about the technology in the abstract and more about what it can make possible for teams operating under real security and compliance pressure. In the Defence Industrial Base, organizations are expected to protect sensitive government data, manage identities across increasingly cloud-native environments and prove that CMMC controls are working in practice. That creates a constant stream of alerts, evidence, access decisions and risk signals that can overwhelm even mature teams. Where AI becomes especially valuable is in helping security teams turn that complexity into action. It can identify patterns faster, surface unusual access behavior, prioritize the risks that matter, support documentation and make continuous monitoring more achievable. For contractors preparing for a CMMC assessment, that kind of speed and visibility can help bridge the gap between security that exists on paper and security that can be demonstrated day to day. But appreciation should not become blind trust. AI is not a substitute for governance, accountability or operational discipline. It can make strong processes faster and more scalable, but it cannot make weak processes defensible on its own. The organizations that will benefit most are the ones using AI to strengthen security processes they already own, not bypass them. For the DIB, that means applying AI responsibly to support identity management, assessment readiness and stronger protection of the data our national security ecosystem depends on," said Heidorn.

Elsewhere in cybersecurity, AI is reshaping how attackers and defenders operate while increasing demand for specialist human input.

Kara Sprague, Chief Executive Officer at HackerOne, said automated tools have changed both the speed and nature of vulnerability discovery.

"AI Appreciation Day tends to celebrate what AI creates. In security, the more interesting story is what it defends against. Attackers adopted AI first. They use it to find and exploit weaknesses faster than any human team could match, and that is exactly why AI has become indispensable to defenders. But AI has not replaced the human security researcher. It has raised the value of one. In the first half of this year, security researchers earned roughly $47 million through our platform, up 25% from the same period a year ago, even as AI made finding vulnerabilities faster than ever. AI can flag a possible vulnerability in seconds. It still takes human expertise to prove that vulnerability is real, understand how an attacker would actually use it, and decide what to fix first. Machines generate the possibilities. People confirm the truth. The organizations getting this right are not choosing between security researchers and AI. They are pairing them. Speed from the machine, judgment from the human. That is what continuous security now demands," said Sprague.

Commercial leaders also see AI agents changing how consumers research products, compare options and complete purchases across digital channels.

Unnikrshnan Kurup, Director of Client Consulting and Strategy at Theorem, said retailers and brands are beginning to redesign journeys around AI-assisted decision-making.

"On this AI Appreciation Day, it is important to remember that AI's value is clearest when it helps people discover things, get educated and make better decisions. In commerce, that is becoming more important as the path to purchase becomes more hybrid across people, content, media and AI agents. Retail media is playing the performance role, video commerce is helping build trust and education, and AI agents are becoming a new way to help consumers make decisions. When we think about the future of AI in commerce, it should sit where it makes everyday decisions easier and more useful. People will keep using AI when the value exchange is clear: it saves me time, it saves me money, it helps me make better decisions and it makes my life easier. A large share of commerce will become agent-assisted, but there will still be a separation between tasks and desires. AI agents can take over routine work like replenishment, comparison, deal hunting, delivery optimization, product filtering and subscription management. Humans will still drive taste, identity, values, gifting, discovery and emotional choices. Rather than replacing people altogether, agents will become a new layer between intent and decision-making. The biggest opportunity is using AI to create convenience without taking away control. AI should help people make decisions they understand and feel confident in, not decisions they cannot explain or reverse," said Kurup.