Skip to content

Adobe opens its software to the MCP protocol and competes with Figma with a prototype tool

Share
Adobe opens its software to the MCP protocol and competes with Figma with a prototype tool

Listen to this article

Read by Anchor

Adobe announced at its annual MAX conference in Miami that it is making the open Model Context Protocol (MCP) available to connect its creative software directly to conversational models such as ChatGPT, Claude and Gemini, shifting its long-standing design tools, including Photoshop, Illustrator, Premiere and After Effects, from closed work environments to programmable endpoints that AI bots can call to invoke functions and control them, alongside the launch of a proactive intelligent agent that runs continuously inside the applications to understand project context and execute fast automated tasks such as background removal and generating video thumbnails.

The most prominent move in the competition front is Adobe’s explicit entry into the space dominated by the Figma platform, by launching a new tool for building digital-project prototypes called Builder within the Firefly app in a private beta, accompanied by equipping Firefly with tools to generate brand-compliant graphics, edit specific parts of images without affecting other elements, and an engine for assembling video clips and sound effects.This shift reveals the company’s aim to settle the early-design phase before the user leaves its cloud umbrella.Conversely, the launch of the MCP protocol gains sharp competitive dimensions after comments by Dylan Field, co-founder of Figma, who said at an event in Bangalore last month that the standard is not ideal and mainly serves the interests of large foundational model labs.

At the technical architecture level, the company has abandoned attempts to monopolise content generation, reinforcing an open-aggregation strategy by making specialised external models from Google, Runway, Eleven Labs, Black Forest Labs, OpenAI, Cling and ByteDance available, and it also introduced two new image-processing models from Topaz Labs, a recent acquisition: Wonder 4 for upscaling and denoising, and Image HDR for converting standard images to high-dynamic-range. Deputy President Dipa Subramaniam explained that Adobe does not seek to restrict what the protocol endpoints expose, but leaves designers to decide how to integrate these capabilities into their daily workflows based on practical experience.Opening automated-control pathways reshapes the relationship between creative production software and external agent systems.

Beyond the overall framework, the updates included direct functional upgrades in the core applications; Photoshop gained a feature for redistributing lighting and generative-style control sliders, Premiere received a colour-grading system aimed at professionals with a tool for removing unwanted video elements, and Lightroom added capabilities for correcting eye-blink closure and adjusting blemishes. These changes, combined with the software-agent layer, make complex processes subject to text commands and automated calls.

For creative production teams and advertising and marketing agencies in the Gulf, Egypt and the Levant, this announcement shifts visual-content management from reliance on time-consuming manual effort to engineered automated production pipelines. The ability of agents and intelligent systems to control design tools directly means local studios will face pressure to restructure their service pricing, as manual production time loses its competitive edge in favour of speed of distribution and the building of software frameworks that link generative models to editing and design programs. The decisive skill for designers and directors in the region becomes managing work context and controlling protocol outputs, while regional tech managers find broader options to connect their internal systems directly to production software without needing dedicated licences or complex intermediary tools.

Don't miss the next story

Subscribe for updates