ServiceNow's new AI agents will happily volunteer for your dullest tasks

ServiceNow has for years used the example of employee onboarding to explain the power of its wares, pointing out that a lot of people around an organization are needed to get new hires on the payroll, registered with HR, equipped with a computer, and assigned appropriate permissions to access applications.

If any of the people responsible for those chores drop the ball, new hires can spend their first days or weeks sitting on their hands. ServiceNow's SaaS-y workflow aims to instead put enrollment chores on relevant people's to-do lists and make sure new hires can quickly get to work with all the requisite tools.

ServiceNow has also spent years banging on about the power of AI. And in its Yokohama release, which debuts today, it's brought onboarding and AI together with agentic tech to automate the jobs required to get new hires up to speed.

As explained to The Register by president and chief operating officer Amit Zavery, the addition of agentic AI means that when news of a new hire is entered into ServiceNow, it will - if customers choose - connect to the applications that a human would today use during the orientation process. Computers could be automatically ordered, bank details entered into payroll apps, and so on. Zavery imagines humans would still check to make sure the agent did the job right. But those humans would be in oversight roles, not stuck doing process work.

Yokohama includes many more agents for diverse tasks. Infosec teams can use "security protocol agents" that, we're told, can unlock accounts if users forget passwords, or deactivate unused accounts.

Other agents will help humans accomplish tasks they've not been trained to achieve.

Zavery used the example of an HR staffer being asked to process a parental leave request for a worker in a country whose laws they don't know. He thinks such tasks can now be handled by an agent, which would log into the relevant systems and do the job - again with human oversight.

Agentic AI is also helping ServiceNow push into CRM by allowing the creation of conversational interfaces customers can use to request assistance or action. Zavery painted a picture of agents enacting those requests when possible, or routing jobs to relevant customer service staff. In his view, CRM therefore becomes more of a customer-facing app and less about internal customer management chores.

The update also adds an AI Agent Orchestrator and AI Agent Studio that together create new agents, link them to perform complex tasks, check they're working as intended, and manage their lifecycle.

Good old generative AI - which, let's face it, is so 2023 now - comes into play in ServiceNow Studio, which allows no-code and low-code development of apps on the company's platform.

Unsurprisingly, it can integrate with AI Agent Studio to build agents.

Another addition is a service observability tool that aggregates info from multiple tools. Of course, AI is involved. ServiceNow reckons that by analyzing pooled observability data it will be able to find problems faster.

As ever, all this stuff is optional because each ServiceNow customer gets their own instance, either in the cloud or on-prem, and can choose the moment at which they upgrade.

The Register asked Zavery if the widespread use of AI in this release means costs will rise, given that a lot of inferencing is required to build and run agents or use generative AI.

He acknowledged that inferencing brings "additional cost" to ServiceNow, but said the sums involved are "insignificant to us." Zavery thinks they'll become even less of a factor due to falling inferencing costs, leaving him happy to be an application vendor rather than a model maker. ®