🤖

Adobe's Agentic AI Innovations

Sep 1, 2025

Summary

  • The session, led by Amit Ahuja (SVP, Adobe Experience Cloud), focused on demystifying Agentic AI, illustrating its real-world impact, and deep-diving into Adobe’s platform innovations.
  • Highlights included a detailed case study with Coca-Cola’s Rapha Abreu and Shekhar Gowda on scaling brand storytelling using AI, demonstrations of Adobe’s new Agent Orchestrator and Agentic reasoning approach, and new agent-driven solutions like Brand Concierge and Adobe Marketing Agent for Microsoft 365 Copilot.
  • Key decisions and advice emphasized maintaining a human-led approach to AI-powered creativity and ensuring brand compliance at scale.
  • The session closed with an invitation to further explore AI advancements at Adobe Summit and hands-on areas like the Agentic zone and Coca-Cola experience.

Action Items

  • (none stated in transcript; see Open Questions for engagement opportunities)

Unpacking Agentic AI and Adobe’s Approach

  • Amit Ahuja set the stage by addressing industry buzz and skepticism about practical AI deployment, emphasizing Adobe’s multi-year journey from predictive AI, through GenAI, to the current Agentic phase.
  • Agentic AI is positioned as the next evolution, defined by interactive, reasoning, and autonomous agents capable of supporting complex tasks and personalized experiences.
  • Adoption statistics showed momentum: 39% of consumers now use AI (e.g., ChatGPT), and 55% of businesses are considering Agentic experiences.
  • Showcased real-world value, e.g., PJ Tours reduced content production time from months to hours; AAA Northeast saw a 156% increase in transactions through AI-generated insights.

Case Study: Coca-Cola’s AI Journey

  • Rapha Abreu (Global Design) and Shekhar Gowda (Marketing Technology) described the need for AI to maintain brand consistency and scale content across 220 markets supporting 2.2 billion daily servings.
  • Key challenge: balancing creative flexibility with strict brand governance and preventing misinterpretation of guidelines.
  • Solution: Project Vision, an AI-enabled dynamic design system powered by Adobe Firefly, which encodes brand style into every design element, ensuring compliance while empowering creative teams.
  • Project Vision is embedded into tools designers use daily, automates brand governance, shortens content production cycles by 10x, and flattens learning curves for global teams.
  • Additional use cases explored include real-time personalization and leveraging Adobe Experience Platform (AEP) for hyper-personalized customer journeys.
  • Advice from Coca-Cola leaders: AI is to enhance—not replace—creativity; start with a clear business problem and clearly define where AI, automation, and human judgment should apply.

Deep Dive: Agentic AI Platform and Reasoning Demo

  • Adobe’s Agentic strategy is structured on three pillars: empowering practitioners, enhancing end-customer experiences, and supporting ecosystem extensibility.
  • Amit clarified Adobe’s definition of an agent: interactive (multimodal), reasoning (contextual), and autonomous (actionable).
  • Shiv Vaithyanathan (VP, Platform Engineering) introduced logical Agentic Reasoning (planning, verification, backtracking), drawing analogies to solving Sudoku puzzles.
  • Demonstration with Akash Maharaj (Lead Data Scientist) showed the Audience Agent in action, guiding a marketer through creating highly targeted audiences using a combination of AI-driven suggestions and human choices.
    • The AI agent handled plan generation, data set selection, model training, verification, backtracking when results fell short, and provided transparency/reporting on outcomes.
    • Emphasis: dynamic, collaborative problem solving where optimal outcomes require both AI and human input.

Next-Gen Brand and Consumer Experiences with Agentic AI

  • Winnie Wu (Director, Product Management) demonstrated Brand Concierge, an agent-powered system for creating deeply personalized, conversational brand experiences.
    • Marketers configure goals, brand guidelines, and content; Brand Concierge surfaces tailored strategies, dynamically adapts content, and delivers personalized experiences for each consumer.
    • The agent continues learning from every interaction, feeding insights back for campaign optimization.
  • Integration scenario: Adobe Marketing Agent within Microsoft 365 Copilot enhances productivity by drafting briefs, recommending audiences and content, managing tasks, and providing campaign analytics within familiar Office workflows.
    • Marketers can execute campaign steps—audience selection, creative inspiration, workflow management, and content editing—directly inside Microsoft Word and Teams.

Ecosystem and Interoperability

  • Adobe reaffirmed its commitment to open Agentic integration for partners, customers, and developers, supporting both Adobe’s and third-party agent orchestration.
  • Demonstrated interoperability with Microsoft 365 Copilot, showing the potential for cross-platform, agent-driven productivity.

Decisions

  • Human-led AI approach prioritized — Both Adobe and Coca-Cola leaders stressed that AI should enhance (not replace) human creativity, with humans retaining final creative and strategic control.
  • Agentic AI design demands transparency, context, and explainability — Ensuring trust, precision, and data safety remain foundational in enterprise applications.

Open Questions / Follow-Ups

  • Opportunity for attendees to explore hands-on demos at the Summit’s Agentic zone and Coca-Cola experience pavilion; participants are encouraged to follow up for deeper dives or direct questions to Adobe product experts.
  • No explicit action items or follow-ups were assigned within the session transcript.