Your browser doesn’t support HTML5 audio
Adobe is offering its customers a new package of end-to-end agentic automations as it opens its annual conference today. The flagship announcement is CX Enterprise, a new agentic platform that builds on the longstanding Adobe Experience Platform and incorporates the agentic orchestration capabilities launched at last year's Summit, as well as adding built-in integration to third-party platforms.
All of this connects into other Adobe components that are also getting agentic makeovers, including the Gen Studio content workflow platform with embedded brand governance, the customer journey optimization for B2B engagement, and a separate analytics layer. In common with other vendors, Adobe's agentic offering has evolved from a set of task- and role-based agents to a multi-purpose agentic assistant, in this case called the CX Enterprise co-worker. Sunil Menon, VP Adobe CX Enterprise Portfolio, explains:
Last year, we talked about Agent Orchestrator and a collection of purpose-built agents that really help you automate activities and tasks. But again, they were in the realm of invoking the agent, performing the activity of the task, and then coming back and then helping you orchestrate across multiple agents.
Where we are now as we think about the evolution, both from a technology standpoint and what we want to bring to market, it's really about this notion of an AI co-worker. Think of this as a super agent that is able to orchestrate multiple agents towards a goal, decompose that into this multi-step plan of execution, and to have the right permissions and guardrails and the scope of expertise that is required. And the other thing being, it is persistent and always on, so if you think about the need for this to take action based on signal or to schedule this.
The agentic co-worker draws on Adobe's own domain expertise and its contextual understanding of the customer. Amit Ahuja, SVP, Products, Adobe Customer Experience Orchestration, says its capabilities will expand over time:
We've been building capabilities around what is a more long-running agent with shared context of the enterprise, and this is one that we're super excited about. The first area that we're really going to showcase at Summit is around some more on the customer engagement side, ie, the journey building, the listening for signal. And how does this agent effectively help humans do the work in that realm of customer engagement?
There's more detail in Adobe's press releases this morning on the launch of CX Enterprise and the new coworker agent as well as a new AI assistant for Firefly, Adobe's creative AI studio. Other releases cover the GenStudio content supply chain offering, a new brand visibility solution and additional integration with the partner ecosystem. We'll have more coverage of Adobe Summit from Madeline Bennett, who is on the ground at the event for diginomica.
My take
I have a sense that Adobe finds itself trapped between three conflicting forces and with no easy way to resolve the conflict. The first of those forces is the relentless progress of agentic AI, bringing end-to-end automation to its domain in ways that require a fundamental rearchitecture of its product set to emphasize cross-functional processes. This is challenging but manageable — the technology is always the easy part. The two other forces are human, and far more intractable.
First of all, there are the demands of its customers, who are set up to operate in particular ways. The new capabilities that AI brings will change those operating models, but there are two separate paths that change can take. The most obvious path is to use automation to improve existing processes, and this is where Adobe seems to be focusing, because that is the course that makes most sense and is most digestible for its customers. But the danger with this is that you risk automating processes that AI will ultimately make redundant.
I believe this is a very real risk in the marketing domain, where for many years, the goal has been to separate the audience into ever-smaller segments in the name of 'personalization at scale'. This was always a workaround, a proxy for actually marketing to each person as an individual, which was seen as logistically and technologically impossible to achieve. But with AI, that goal of individualization at scale now becomes achievable, and the old model of marketing to ever-smaller segments becomes redundant. But organizations locked into that old model are so focused on continuing to perfect it that they aren't even thinking about how it might be replaced with something new. This is the classic bind for incumbent vendors, and it seems to me that Adobe is locked into continuing to servicing that redundant goal.
Then there are the users, who are not even up-to-speed with ways of working that would allow any form of automation to work efficiently, whichever model it follows. In recent years, Adobe has been championing the notion of the content supply chain, in which assets are shared and teams progress their work through digital processes, with brand guidelines automatically enforced. Making these processes digital are a pre-requisite for applying AI to them. But many design and marketing professionals are wary of sharing their work in such a transparent fashion, fearing a loss of control and autonomy. Overcoming this resistance requires careful change management, and the additional disruption that AI will bring to work routines makes this challenge even harder.
These twin forces — or perhaps rather than a force, it's more accurate to describe them as inertia — provide stubborn resistance to the way forward enabled by the technology. It's the curse of incumbency that Adobe has to deal with, and to a large extent placate, these factors, while new rivals aren't held back in the same way. All of this doesn't mean that Adobe can't succeed in the AI era — there are huge revenues to be earned from continuing to serve its existing customers and users, even while they pursue less optimal paths than would be in their better interests. Meanwhile, Adobe is making heroic efforts to adapt its platform to the demands of AI. But satisfying the more conservative desires of its customers and users while at the same time building an AI-native architecture that's fit for the future would strain the capacity of any organization, and Adobe is no exception. The next year or two will be decisive in showing whether it's up to the task.