Exploring new possibilities with connected, AI-powered solutions.
We’re standing at the edge of a major technological shift: one where intelligent systems won’t just understand and respond, but take meaningful action to support and enhance our work.
We’re entering the next phase of AI—agentic AI—where systems evolve from helpful assistants to context-aware, goal-driven agents that reason and action. These agents will operate with greater awareness, help make decisions in real time, and connect seamlessly across workflows—enabling them to act as orchestrators of the AI tools and other resources professionals need to achieve their project goals.
While society isn’t fully there yet, the path forward is rapidly taking shape. It not only introduces new opportunities and possibilities, but also a renewed standard for security and trust.
To explore what this shift means—for our customers, our company, and the entire developer community—Autodesk Chief Trust Officer Sebastian Goodwin and I sat down to chat about what it will take to responsibly usher in this next wave of AI:
Raji: Agentic AI represents a fundamental shift in how AI systems function. We’re moving beyond the old point-and-click model to a new era of “describe and do,” where your intent becomes action. Instead of waiting for human input at every turn, these systems will be capable of initiating actions and making decisions based on and within context defined by humans—as well as interacting with tools, data, and environments to do so.
To be clear, these agentic workflows are something our world is still working towards—but it’s an exciting time because we’re seeing the groundwork rapidly laid. Technologies like Model Context Protocol (MCP) servers are currently serving as intelligent USB connectors of sorts, enabling AI models and agents to find data, APIs, and skills on the fly. They’re providing the foundational, standardized infrastructure that will enable these models to access a broader set of actions, which can be dynamically combined to deliver meaningful outcomes. They can learn, self-correct, and improve in real time.
So, as we stand on the edge of this transformation, now is the time to explore, experiment, and embrace this technology with care—to ensure the agentic AI systems our customers and the developer ecosystem have access to are not only powerful, open, and extensible, but also responsible, reliable, and trustworthy.
Sebastian: This is one of the most important questions we can ask right now. Greater autonomy means an even greater need for trust and accountability.
As we move towards a more agentic world, trust can’t be an afterthought. It’s not just about protecting data or endpoints. It’s about embedding safety, governance, and oversight directly into the AI’s operating environment to ensure these agents act within their intended bounds.
It’s also essential to have technical controls in place to keep contextual memory and agent interactions secure and reliable. This includes AI-specific safeguards—like setting expiration rules for shared context or validating data formats to block malicious inputs. There’s also a big opportunity to use agents themselves to enforce trust. For example, Context Integrity Agents or Policy Enforcement Agents could be introduced to help monitor and protect the environment in real time—not just individual components. As these systems become more connected, trust will need to be managed at the ecosystem level, potentially through a shared, coordinated Trust layer.
But let’s take a closer look at the MCPs gaining rapid adoption across industries as we speak. The power of these MCPs lies in their ability to trigger action. But what if that action was unreliable, untrustworthy—or worse, malicious?
MCPs are currently serving as bridges between AI models and real-world actions, systems, and data, which amplifies both the benefits and risks associated with AI. So, if an AI model is inaccurate or biased, or if a database isn’t secure, that bridge becomes a point of failure, leakage, or attack. As more AI agents are granted permission to operate in the real world, the risks increase if those connections aren’t built with care.
From an Autodesk perspective, this is what we’re thinking about—not just in the current rise of MCPs, but also in shaping the future of agentic AI. As always, we’re deeply committed to building and using AI responsibly, guided by our Trust principles for AI. These principles help ensure our work stays grounded in transparency, accountability, reliability, and security. Our goal here is simple: to help our customers and the broader developer community tap into the potential of this emerging technology, safely and confidently.
Raji: Preparation starts long before deployment. We’re investing time and talent today with the goal of unlocking greater capacity for our customers as they move toward a more agentic future.
We’re using our expertise in design and make to explore early use cases that tackle some of our customers’ most pressing industry challenges around data, intellectual property, and collaboration.
We’re shaping interoperability standards, automating manual workflows to drive productivity, and stress-testing how these agents can help humans create tangible outputs that perform in the physical world.
And, we’re working to ensure the Autodesk AI-powered solutions our customers have access to not only deliver intelligent support, but do so with a deep understanding of their respective industries.
We’re also being clear-eyed about the timeline. The shift to agentic AI isn’t a distant horizon. Things are moving fast and we’re helping our customers get ready for what’s next: more connected workflows, faster paths from idea to impact, and a future where creative and operational velocity increases dramatically.
Additionally, above all, we’re showing up and engaging as a trusted partner. Autodesk has spent the last 40+ years guiding our design and make customers through digital transformation. We’ve also spent the last decade investing in AI, APIs, and granular data. Agentic AI is a natural extension of that journey, and we’re committed to ensuring every step forward is grounded in trust. The possibilities are enormous, and so is our responsibility to help shape this future with purpose.
As we enter this new chapter together, our mission is clear: to help our customers navigate this important technological shift with confidence and ensure their use of Autodesk AI systems remains secure, transparent, and trustworthy.