Think about the last time you got genuinely lost. Not turned-around lost — actually, no-idea-where-I-am lost. If you're under fifty, it probably hasn't happened in years. Not because you got better at navigation. Because the tool got so good, so accessible, and so woven into daily life that getting lost became almost impossible. That's not a feature update. That's a category shift — the moment a technology stops being something you use and becomes something you rely on without thinking.
I've been watching LEAP approach that same threshold for a while now. And with this release, I think we've crossed it.
LEAP started as a pattern recognition engine — a way to look across your incident history, identify recurring problems, and surface the automation opportunities hiding inside them. The underlying logic was always about compound learning: every incident your team resolves teaches the system something, and the system uses that to make the next response faster and more precise. The potential was clear. But like GPS before it was in everyone's pocket, the full value wasn't equally accessible. Getting there required specific licensing, specific configurations, and a level of AI maturity that not every organization had reached yet.
That changes now. This release opens LEAP's AI capabilities to all customers, with the ability to choose the LLM that fits your environment — ServiceNow's native model, OpenAI, Gemini, or Anthropic — at no additional cost. The intelligence LEAP has been building in your environment is no longer behind a door that only some customers can open.
When Your Own Data Isn't Enough
Here's a situation every operations team knows: an incident surfaces on a system your team hasn't seen fail before. Your internal knowledge base is thin. Your work notes don't cover it. LEAP, up to now, could only work with what was already inside your environment. This release adds web search — via Gemini or Perplexity — as a third source alongside internal work notes and KB articles. When the internal signal is weak, LEAP reaches further. The result is more complete resolution steps in the situations where your team needs guidance most and has the least to work from.
The practical outcome: fewer incidents where operators are on their own, fewer escalations to subject matter experts for issues that have well-documented solutions elsewhere, and faster resolution on the long tail of incidents that fall outside your team's direct experience.
When 300 Incidents Is Actually Five Different Problems
Large incident clusters have always been a friction point in LEAP. The system identifies a pattern — say, 300 incidents with a common thread — and generates resolution guidance for the group. But operations leaders know that a cluster that large almost always contains several distinct sub-problems. Generic guidance for 300 incidents is not particularly useful guidance for any of them.
Refined opportunity grouping addresses this directly. LEAP now breaks large clusters into focused sub-groups based on shared sub-categories, and generates tailored resolution steps for each. The difference in practice is significant: instead of guidance an operator has to interpret and adapt before acting, they get specific steps they can apply immediately. For a team managing dozens of open opportunities at any given time, that specificity translates directly into faster closures and more automation opportunities that actually get implemented rather than deprioritized.
The Knowledge That Used to Disappear
Every time LEAP generates resolution steps that an operator successfully applies, something valuable is created. Until now, that value largely evaporated when the incident closed — useful in the moment, but not captured anywhere a future operator could find it.
KB Article generation from Resolution Steps closes that loop. Operators can now promote LEAP's resolution output directly into the knowledge base, turning a single incident's resolution into a reusable asset for every operator who encounters that pattern going forward. This is compound learning made tangible: the system gets smarter, but so does the organization around it. Over time, the knowledge base stops being a static library and starts reflecting the actual solutions your team has validated in production.
Guidance at the Moment It Matters
The enhanced SOW integration takes the KB article capability one step further. When a new incident opens, LEAP-generated articles for related patterns surface directly in the Service Operations Workspace — without the operator having to search for them. The guidance arrives at the moment the operator is trying to decide what to do, not after they've spent time looking for it. For high-pressure, time-sensitive incidents, that difference in timing has a measurable effect on MTTR.
For teams earlier in their LEAP journey, simplified setup consolidates configuration — savings parameters, priority weighting, grouping criteria — into a single dashboard with inline guidance. The goal is the same as everywhere else in this release: remove the friction between a team and the value they're trying to reach.
One More Thing Worth Knowing
For customers running Cloud Accelerate and Cloud Governance Suite, this release adds integrated spend visibility from Cloud Cost Management directly into Cloud Workspace. If you hold a CCM entitlement, account-level spend trends, monthly variance, and budget tracking now live alongside inventory, ownership, and security data in a single view. Cloud leaders have been asking for this — not because the spend data doesn't exist elsewhere, but because decisions made without the full governance picture tend to be incomplete ones.
What This All Adds Up To
The destination hasn't changed: zero service outages, reached through a progression from accelerated response to self-healing operations to fully preventative IT. Compound learning is the engine that makes that progression possible — each incident makes the next one easier to handle, and eventually, easier to prevent.
What this release does is remove the barriers that were slowing that progression down. Broader AI access means more organizations can start accumulating that learning now. Better resolution guidance means more incidents get closed cleanly the first time. KB generation means the knowledge sticks. SOW integration means operators can act on it faster.
Your GPS didn't get smarter because the maps improved. It got smarter because it learned from everywhere everyone was going, all the time, and used that to get better at guiding each individual driver. LEAP works the same way. The more your organization uses it, the more precisely it serves you — and now, nothing is standing in the way of that flywheel starting to turn.
