AI shifts from hype to resilient business infrastructure

Technology leaders are using AI Appreciation Day to highlight a shift towards more pragmatic, resilient uses of artificial intelligence in business. Executives at Commvault and Alphatax say AI is moving from hype to infrastructure and deeper automation.

Vendors and customers are reassessing AI strategies as boards scrutinise cost, return on investment, and operational risk. Industry figures say the conversation has moved beyond experimentation and headline features. AI now underpins workflows, shapes security assumptions, and reshapes white-collar roles in fields such as tax and finance.

At data protection specialist Commvault, Field CTO Gareth Russell said AI sits closer to the core of enterprise technology stacks than many organisations realise. Firms must align AI adoption with cyber resilience, data protection, and identity controls rather than treat it as a separate innovation layer, he argued.

"Not long ago, AI felt like a competitive advantage. Today, it is rapidly becoming infrastructure.

Russell contrasted earlier technology shifts with the current wave of automation driven by AI agents. He cited Anthropic's Mythos demonstration as an example of how AI can compress the time between vulnerability discovery and exploitation from weeks to minutes, changing the economics of cyber risk more than the nature of the threat itself.

Concerns about bias and hallucinations remain, but the debate is widening to include the integrity of the data, identities, and recovery mechanisms surrounding AI tools, he said.

"Every generation has a technology that rewrites the rules instead of simply improving the game. Cloud changed where applications lived. Mobile changed where work happened. AI is accelerating the pace of business," Russell said.

"Take Anthropic's recent Mythos demonstration. The headline was not that AI could identify software vulnerabilities. Security teams have been managing vulnerabilities for years. More significant was the reduction in the time between discovery and exploitation, from weeks to minutes. AI is not creating an entirely new problem. It is fundamentally changing the economics of an existing one," he said.

"That is why AI Appreciation Day should be about more than celebrating innovation. It should also be about appreciating the responsibility that comes with it.

"Over the past two years, much of the conversation around AI has focused on hallucinations, bias, and misinformation. Those conversations remain important. But as AI becomes more autonomous, the discussion is expanding. It is no longer just about whether AI produces trustworthy answers. It is whether we can trust the ecosystem around it.

"Can we trust the data feeding it? Can we verify the identities interacting with it? Can we recover the systems it depends on after an attack? Can we understand and govern the decisions made by increasingly autonomous AI agents?

"Those are the questions organisations will need to answer as AI becomes embedded in everyday business.

"Innovation now depends on resilience. AI is only as trustworthy as the data, identities, and infrastructure behind it. If any of those are compromised, trust erodes quickly, along with the productivity AI promises.

"At Commvault, we've long believed responsible AI and cyber resilience go hand in hand. Responsible AI is about more than governance and ethics. It is about building AI that is transparent, accountable, secure, and supported by data organisations can trust and recover when the unexpected happens. Trust does not come from AI alone. It comes from the resilience of the entire ecosystem that supports it.

"The next evolution of work is already beginning to take shape. Employees will not just use AI. They will increasingly work alongside autonomous agents that research, reason, automate, and act on their behalf. Soon, bringing your own AI agent to work may feel as natural as bringing your own device once did.

"When that happens, the competitive advantage will not come from having more AI. It will come from creating an environment where people and AI agents can operate together with confidence. That requires resilient data. Trusted identities. Transparent governance. Recoverability by design. In other words, the foundations that allow innovation to scale without sacrificing trust.

"Perhaps that is the real lesson of AI Appreciation Day. Appreciate AI for what it can achieve today. But prepare for what comes next.

"Because the future of AI will not be defined by intelligence alone. It will be defined by trust. And in the age of the agentic enterprise, resilience will be the foundation that makes that trust possible," Russell said.

In tax technology, Alphatax executives say AI is beginning to change both workflow automation and the experience of professionals working under heavy regulatory pressure. They argue that agentic systems can handle more of the complexity and exception handling that has historically limited automation projects.

Chief Innovation Officer Russell Gammon said AI marks a break from the constraints of robotic process automation in tax functions.

"There's a lot of discussion about what AI can do for tax professionals, such as analysing trial balances, speeding up regulatory research, and generating more tailored client queries. But one of its biggest contributions is less obvious: it is fundamentally changing what automation is capable of," Gammon said.

"For years, robotic process automation was the standard approach to automating repetitive tax processes. Compared with spreadsheets and other legacy macros, it represented a significant step forward.

"But AI is now leaving robotic process automation in the dust. It enables automations to be built and maintained more quickly and at a lower cost than traditional robotic process automation projects. If a process changes, updating it is often as simple as changing a prompt rather than rebuilding an automation workflow. More importantly, agentic AI is extending automation beyond rigid, predefined rules. Instead of stopping when something unexpected happens, AI can interpret context and investigate exceptions to find alternative routes to the final goal.

"In practice, this means organisations no longer need to think purely in terms of automating individual tasks. They can begin to automate larger parts of workflows that previously depended on manual intervention whenever an exception occurred. For tax professionals, that means spending less time keeping processes moving and more time applying their expertise where it matters most: making tax decisions.

"Robotic process automation has played an important role in the evolution of tax technology, just as Excel macros did before it. But every generation of technology has its moment. Now AI is taking automation into its next phase by making it more intelligent and far better equipped to deal with the complexity that has always existed in tax," he said.

Alphatax Chief Executive Officer Bruce Martin highlighted AI's impact on workload and wellbeing in tax and finance teams. Practitioners face rising data volumes, intricate rules, and strict deadlines, and automation is beginning to change the nature of daily work, he said.

"In industries like tax and finance, where professionals often manage high volumes of data, complex regulations, and intense reporting deadlines, AI can help reduce workloads and remove some of the mundanity that contributes to stress and burnout. By automating repetitive tasks such as data entry, employees are freed up to focus on more strategic work. That shift can have a significant impact on motivation, engagement, and overall job satisfaction," Martin said.

"We are also beginning to see AI play a more supportive role in day-to-day working life. AI assistants can act as technical sparring partners, helping employees test ideas, brainstorm solutions, and work through challenges in a collaborative way. In many cases, these tools provide instant feedback and positive reinforcement, which can help boost confidence and reduce the isolation people sometimes feel when working independently.

"This is particularly relevant in today's hybrid and remote working environments. While flexible working has delivered enormous benefits for work-life balance, it can also leave some employees feeling disconnected. AI tools, when personalised with the right context about an individual's role and working style, can recreate some of the interaction people would naturally receive in an office environment," he said.