ServiceNow's new bid to control enterprise AI

Welcome to Runtime! Today: ServiceNow introduces a new version of its Control Tower agentic AI service, Anthropic and OpenAI are getting into the consulting business, and the latest funding rounds in enterprise tech.

Please forward this email to a friend or colleague! If it was forwarded to you, sign up here to get Runtime each week, and if you value independent enterprise tech journalism, click the button below and become a Runtime supporter today.

Become a Runtime Supporter

Buzz the tower

Last year at ServiceNow's Knowledge event CEO Bill McDermott promised that "autonomous agentic AI" was right around the corner, and that his company's services offered the IT departments of the world the best chance to harness that shift. While the agentic part is starting to round into form a year later, most companies aren't convinced the agents are ready to fully take over their mission-critical workflows.

ServiceNow introduced several new services at Knowledge 2026 that it hopes will convince customers to put more trust in the machine, including an updated version of the Control Tower AI service unveiled at last year's event and an expansion of the Autonomous Workforce agents launched in February. "We're learning that the real competitive differentiator is the orchestration surrounding the models," McDermott said in his opening keynote Tuesday.

Control Tower can now find AI applications running across third-party companies like the Big Three cloud providers and SaaS stalwarts like SAP and Workday thanks to 30 new integrations.
ServiceNow also built Traceloop's observability technology into the platform after acquiring the company in March, which allows customers to put tighter reins on AI agents and better understand how they behave.
Technology from another recent acquisition, Veza, improves the security posture of the platform by giving customers a "kill switch" to hit when a problem is detected, according to the company.
"What we launched last year gave customers a governance layer, but what we're shipping this year goes significantly deeper, evolving from visibility and management into a full enterprise AI command center,” ServiceNow's Nenshad Bardoliwalla told The Register.

While Control Tower wants to help companies govern their own AI agents, Autonomous Workforce is a set of pre-built agents designed for ServiceNow's core product areas, like IT service management, HR services, and its budding entry into the CRM market. "Six out of 10 companies are actually using agentic AI, but only one out of 10 have built anything autonomous," McDermott said.

IT managers will be able to deploy new agents that ServiceNow said can handle SRE tasks like triaging an incident, as well as asset-management robots that can help companies understand how well they're utilizing their computing resources.
Sales leaders can ask an agent to qualify leads and prepare quotes for prospective customers based on transcripts of meetings with those customers.
And CISOs can deploy agents that ServiceNow promised will help SOC teams deal with unfolding incidents and evaluate potential software supply chain threats.

A lot has changed since last year's Knowledge event, including the release of new reasoning models from Anthropic and OpenAI last November that turned agentic AI apps from experiments into reality and the collapse of SaaS stock prices (ServiceNow is down 53% since last May) on worries that AI tools will let companies simply build their own enterprise apps (were that so simple). However, one key thing hasn't, and that's the push from every enterprise software company to become the control plane for agentic AI.

"Nobody has ever seen their business this way before, because nobody had ever built a platform that could enable them to see things differently," McDermott said, making the case for ServiceNow's approach.
But the company is in the middle of a crowded race to position itself at the center of enterprise tech AI strategies, and we're still in the process of figuring out what that future looks like.
"Customers should consider the combination of workflows, AI, data, governance, and security as they deploy AI capabilities. No one model can do it all," IDC's Stephen Elliot told CIO.

Deployed even forwarder

Big companies have long turned to services organizations to help them integrate emerging technologies into their existing workflows; at one point a few years ago, early in the AI boom, it seemed like Accenture was making more money from AI than anybody else. In response, Anthropic and OpenAI invested in what they called "foward-deployed engineers" (read: consultants) to help customers build AI applications and agents, and now they've decided to formalize that practice.

Anthropic launched a new, unnamed services organization Monday in partnership with Blackstone, Hellman & Friedman, and Goldman Sachs that "will work with mid-sized companies across sectors to bring Claude into their most important operations," it said in a blog post. Right around the time of that announcement, Bloomberg reported that OpenAI is gearing up to launch a similar venture called The Deployment Company backed by $4 billion in investment from TPG, Brookfield Asset Management, Advent and Bain Capital.

Enterprise AI is a new discipline, and smaller companies that just started to feel comfortable with the notion of cloud computing are in for years of work learning how to use these tools safely and effectively without breaking their core applications. "Companies from community banks to mid-sized manufacturers and regional health systems stand to gain from AI, but lack the in-house resources to build and run frontier deployments," Anthropic said in its post.

Enterprise funding

Sierra raised $950 million in Series E funding, valuing the agentic customer service company founded by ex-Salesforce co-CEO Bret Taylor at more than $15 billion.

Blitzy landed $200 million in new funding for its software development platform, which helps companies automate the work beyond coding such as testing, deployment, and documentation.

Portland's own Panthalassa scored $140 million in Series B funding for its aquatic computing nodes, which generate power from ocean waves to run AI inference jobs over a satellite network.

Netomi raised $110 million in Series C funding for its customer experience technology, which uses AI agents to help companies understand how customers are using and reacting to their products on the internet.

DeepInfra landed $107 million in Series B funding for its AI inference computing platform, which currently runs across eight U.S. data centers.

CopilotKit scored $27 million in Series A funding for its AI development platform, which helps customers build AI applications around the AG-UI protocol that defines how agents interact with user interfaces.

The Runtime roundup

Google, Microsoft, and xAI agreed to allow the U.S. Center for AI Standards and Innovation to review new AI models before they are released, joining Anthropic and OpenAI in a program established in 2024.

AMD's data-center revenue surged 57% during the last quarter, easily beating analyst estimates as AI chips continue to fly off the shelves.

Amazon employees will be allowed to use AI coding agents from Anthropic and OpenAI, according to Business Insider, after somebody finally decided that was probably enough dog food.

A Microsoft developer apologized for adding a "Co-authored by Copilot" message to code commits written using the VS Code editor even if users had disabled all the GitHub Copilot features in the popular tool.

Thanks for reading — see you Thursday!