Thoughtworks launches Agent/works for AI governance

Thoughtworks has launched Agent/works, a platform for governing and running enterprise AI agents across any cloud. It provides a single control plane and a governed runtime for AI agents.

The launch comes as companies move from experimenting with AI tools to deploying autonomous systems that can access data, use software tools, and carry out workflows with less direct human input. That shift is creating new pressure on security, compliance, and finance teams as AI systems spread across organisations.

Agent/works is designed to give companies a single registry for agents deployed across their technology estate, with oversight of governance, performance, and AI spending. It is intended to sit beneath custom agent-based applications built by enterprise product teams, rather than operate as a standalone application.

Thoughtworks framed the product around what it calls agent sprawl: organisations losing visibility into how many AI agents are in use, what data they can access, and what resources they consume. It also linked the issue to the spread of AI-assisted software development, which is increasing the volume of code and automated workflows entering production systems.

Industry research cited by Thoughtworks points to both adoption and risk. According to Sonar's 2026 State of Code Developer Survey, developers reported that 42% of committed code is now AI-generated or AI-assisted. Research from AppSec Santa found that 25% of AI-generated code samples contained critical security vulnerabilities.

Governance focus

Agent/works includes controls aimed at managing how AI agents behave before and during execution. The system checks workflow paths before an agent runs to confirm that at least one compliant route exists from end to end.

It also applies permissions tailored to agents rather than human users. Those permissions are capability-based, scope-bound, and time-limited, and can narrow automatically when an agent interacts with more sensitive systems or data.

The runtime is built to support different types of agents, including autonomous workflow agents and interactive coding agents. Policies can adapt during execution, while the platform can connect with different models, tools, cloud services, and third-party agents through standard interfaces.

A central registry sits at the core of the platform, providing visibility into agents, models, tools, and policies, alongside evaluations, usage analytics, and cost controls.

Shayan Mohanty, Chief Data and AI Officer at Thoughtworks, linked the launch to the operational demands of large-scale autonomous systems. "When anyone can generate software with a text prompt, AI governance isn't a checklist you bolt on after the fact. It's a foundational requirement for operating autonomous systems at scale," Mohanty said.

He also tied the issue to cost management as AI agents become more common inside enterprises. "Every AI-powered workflow now carries an operating cost. The challenge for enterprises is no longer just how to build agents, but how to govern the resources they consume. Without runtime controls, costs can scale as quickly as the agents themselves. Agent/works gives organisations visibility and governance across their agent ecosystem, helping them control spend, manage risk, and scale AI with confidence. Governance done this way stops being a brake and becomes the engine that lets teams move fast, safely, and at scale," he said.

Databricks tie-up

Thoughtworks also highlighted its work with Databricks on enterprise AI governance and agentic systems. The relationship centres on helping companies apply governance models used for enterprise data to agent workflows.

David Nasi, Director of Product Management, AI and Agentic Platform at Databricks, said that approach is gaining traction among early adopters. "The organisations moving fastest with agents are extending the same governance models already used for enterprise data across agent workflows themselves," Nasi said.

He added: "This shift matters because it turns governance from a bottleneck into an enabler, making it much easier to scale autonomous systems safely across the entire enterprise."

Thoughtworks said its agentic development offering, AI/works, already runs on Agent/works, which it presented as evidence that the platform is being used in production. It positioned Agent/works as a foundational software layer for enterprises trying to manage the growing complexity of autonomous AI systems while maintaining oversight of risk, compliance, and spending.