Fast followers will fall behind in the AI race, warns ServiceNow

Simplifying interactions

Orica’s initial use case is simplifying interactions with its multiple HR systems, according to Leo Luk, the company’s global process strategy and enablement manager. Orica needed a product that combined process orchestration with multilingual natural language processing and accessibility via Microsoft Teams. It settled on EmployeeWorks after evaluating several alternatives.

The initial focus was on high-frequency tasks, such as leave requests, or processes that involve assembling multiple data fields – sometimes as many as 30 – from various applications.

EmployeeWorks went live at Orica’s North American operations the night before the recent ServiceNow conference, with Luk reporting that the regional HR team had already used it successfully to create a new job position.

Putting AI on top of a single system isn’t enough, said Andrew. Rather, businesses need AI workflows that span across systems, which is exactly what EmployeeWorks provides.

Another recent ServiceNow introduction is AI Control Tower. The product goes beyond monitoring by actively taking action, and is designed to work seamlessly with any AI model and application. “AI was popping up everywhere at Orica,” said Irene Klymenko, the firm’s senior manager of AI portfolio and delivery.

As expected for an explosives manufacturer, there were strict concerns around risk, governance and responsible use. However, the company was determined not to unnecessarily delay innovation. “We can’t scale what we can’t see,” she said.

While Excel and SharePoint were initially used to track AI projects, this approach quickly proved inadequate.

ServiceNow’s AI Control Tower provides a way to action policies and frameworks, pointed out Adrienne Maxted, ServiceNow risk and security partner at Deloitte. She added that it helps with risk assessments, provides asset visibility and supports value realisation by directly connecting AI investments to business outcomes.

This approach to governance has streamlined Orica’s processes, said Klymenko, largely by enabling a move away from manual spreadsheets.

Asked what they wished they had known earlier in their AI journeys, the panellists offered a variety of insights.

Luk pointed to the need to deeply understand the value AI agents provide, and how that value can be measured. Some AI projects at Orica were dropped simply because they were “not moving the needle enough”, he revealed.

Klymenko recommended starting with governance from day one, noting that Orica’s executive team championed this approach.

Maxted highlighted the critical need for stakeholder trust, suggesting early engagement with an organisation’s security and data teams to foster confidence.

Doug Schairer, technology and transformation partner at Deloitte, warned that if organisations fail to build AI into existing workflows, they risk sparking another round of shadow IT, similar to previous waves of major technological change. While shadow IT may get things done quickly from an employee’s perspective, the absence of governance inevitably leads to suboptimal spending, security vulnerabilities and inconsistencies across departments.

The autonomous workforce

Andrew said the autonomous workforce is rapidly approaching, “starting in our core sweet spot” of IT support.

He suggested that routine IT tasks are generally mundane, and typically handed to junior staff – not to develop their skills, but simply because no one else wants to do them.

To address this, the first of ServiceNow’s fully autonomous agents is the L1 Service Desk AI Specialist, which can take care of commonplace tasks such as password resets and network troubleshooting.

“We can turn it on for any customer really quickly,” said Andrew, noting that it “builds trust in agentic employees”.

More than 90% of ServiceNow employees’ own IT requests are already being handled by this agent. It will soon be joined by other IT specialisms, as well as autonomous agents for functions such as HR, procurement and finance.

ServiceNow is seeing widespread interest in its AI capabilities across all industries, said Andrew. Most – if not all – company boards are actively pushing AI initiatives. “They’re all leaning into the conversation,” he added, adding that the idea of ServiceNow as a platform of action resonates widely.

Ultimately, said Andrew, organisations that get the foundations right – particularly when it comes to governance – will be the ones that deploy AI the fastest and most successfully.