ServiceNow's $136B Multi-Model Chess Game: Anthropic Deal Drops One Week After OpenAI

Remember when picking a database meant PostgreSQL or MySQL? ServiceNow apparently didn't get that memo.

The $136.77 billion workflow automation giant just inked a multi-year partnership with Anthropic, making Claude their "preferred and default model" for AI-driven workflows. Plot twist: this happened exactly one week after they announced a similar deal with OpenAI.

Translation: We're playing the field.

The Numbers Don't Lie (But They're Expensive)

ServiceNow isn't exactly hurting for cash. Their Q4 2025 results show:

• $3.57 billion in revenue (up 20.5%)
• $12.85 billion in remaining performance obligations
• 603 customers spending over $5 million annually
• Full-year revenue hit $13.28 billion

With numbers like these, they can afford to date multiple AI models simultaneously. But should they?

The Anthropic deal goes deep. Claude Code (Anthropic's coding assistant) is rolling out to all 29,000 ServiceNow employees. Claude becomes the default brain behind their ServiceNow Build Agent - the tool that lets enterprises create "agentic workflows" without needing a computer science degree.

Healthcare Gets the VIP Treatment

Here's where it gets interesting. ServiceNow is specifically targeting healthcare and life sciences with Claude integration. They're promising to streamline claim authorization from days to hours.

That's not just efficiency porn - that's real money. Healthcare administrative costs in the US hit $500+ billion annually. Shaving days off claim processing could save hospitals millions.

Corporate speak aside, McDermott has a point. ServiceNow isn't building chatbots. They're embedding AI directly into the mundane-but-critical workflows that run modern enterprises:

• IT service management
• HR onboarding
• Supply chain optimization
• Security incident response

Hot Take: The Multi-Model Strategy Is Brilliant (And Risky)

Most companies are going all-in on one AI provider. ServiceNow is doing the opposite - and I think they're onto something.

The brilliant part: Different models excel at different tasks. GPT-4 might crush creative writing while Claude handles code analysis better. Having both means ServiceNow can route tasks to the optimal AI brain.

The risky part: Integration complexity multiplies. Instead of mastering one AI model, ServiceNow's engineers now need to become experts in model orchestration. That's hard.

Plus, both OpenAI and Anthropic are watching each other through ServiceNow's implementation. Competitive dynamics could get weird fast.

The 50% Promise

ServiceNow claims this partnership will cut customer implementation time by 50% - from initial sales conversations to fully autonomous deployment. That's a bold promise for enterprise software, where "quick deployment" usually means "under six months."

If they deliver, it could be game-changing. Enterprise AI adoption is still hampered by complexity and integration nightmares. ServiceNow is betting they can abstract that pain away.

The Real Winner: Enterprise Choice

Forget the corporate messaging. The real story here is optionality. ServiceNow customers won't be locked into one AI provider's roadmap, pricing model, or inevitable outages.

When GPT-5 drops or Claude gets a major update, ServiceNow can route traffic accordingly. When one provider has capacity issues, they flip to the other.

That's not just smart product strategy - it's insurance against the rapidly evolving AI landscape.

ServiceNow stock is trading near historical lows despite strong fundamentals. Maybe the market doesn't understand their multi-model bet yet. Or maybe two AI partnerships in one week looks like expensive hedging rather than strategic brilliance.

Time will tell. But in a world where AI model performance shifts monthly, having multiple dance partners seems pretty smart.