From Agencies to Agents: The Shift Brands Can’t Ignore

The Next Level Of AI Is Agentic–Here's Where To Start

Today marks a significant milestone in AI’s evolution: It’s the two-year mark of when Co-founder (and now Chief AI Officer) Wesley ter Haar declared that AI would reshape advertising’s economics. And he was right: it’s completely transforming the game by elevating creativity, personalizing at scale and revolutionizing the way brands connect with people.

The latest stage of that transformation is agentic AI’s emergence as a major focus for brands, sparking conversations across industries eager to explore its potential. Unlike LLMs, which respond to input, or AI agents, which execute tasks with human oversight, agentic AI takes things further by solving problems and pursuing goals with minimal human intervention.

Agentic AI is already transforming how brands reduce friction and unlock value. For instance, some are using AI agents to optimize ad campaigns in real time, tailoring messaging to audience behavior as it happens. Others are integrating these systems to improve marketing workflows, reducing costs and redirecting teams toward more strategic, creative initiatives. By automating repetitive tasks and enhancing customer engagement, agentic AI enables brands to focus on delivering better outcomes—like faster service, more personalized experiences and improved business results.

Of course, as with any powerful innovation, agentic AI comes with its own challenges—data privacy, security, complexity and more. To help navigate this space, we’ve outlined a few key considerations in our latest Labs Report about the topic.

A few things to do:

Start small and scale up as needed
Prioritize data privacy and security
Maintain human oversight

A few things to watch out for:

Rushing into full automation or overly rely on it
Neglecting data quality and bias
Skipping testing and refining AI’s behavior

Explore the Latest AI Agents from Google and OpenAI

Does the info above have you curious to try using an AI agent of your own? You’re in luck, because Google and OpenAI have both recently released ones that are ready for experimentation.

Gemini Deep Research: Google has released Deep Research, an agentic feature for Gemini Advanced subscribers that autonomously researches complex topics and provides reports based on the user’s query, saving them time on research tasks. But more importantly, the feature showcases Gemini’s growing ability to act independently on the user’s behalf.

OpenAI’s Operator: Operator is an AI agent that uses a browser to perform web tasks for users. Powered by the Computer-Using Agent (CUA) model, it’s able to interact with websites to execute tasks and self-correct errors. It’s currently available to US Pro users. Operator demonstrates a significant step towards more capable and helpful AI assistants. Its ability to interact directly with web interfaces expands AI's utility, automating tasks and improving efficiency across various domains. Though impressive, it still requires human oversight.

We go more into detail about our vision for agents like Operator and its potential to “deliver a ‘big idea’ for every audience, perhaps even every customer” in this recent The Drum article. And if you’re curious to learn about even more recent releases of AI tools, check out our latest Creative Technology Inspo deck.

How Brands Are Overcoming Common AI Obstacles to Reshape Media and Entertainment

As generative AI continues to transform industries, the media and entertainment sector is finding innovative ways to harness its potential. Like many companies in other industries, brands in this space face mounting pressure to deliver more personalized, immersive and scalable digital experiences—often with fewer resources. Generative AI offers a powerful solution, but implementation is not without its challenges.

Forrester’s research, showcased in this AWS on-demand video panel, reveals several struggles brands are facing right now with GenAI. For example, many firms are struggling to implement GenAI effectively and lack the necessary data assets. Some lack appropriate training data. Others experience employee resistance. Legacy systems and poor data management impede organizational objectives when it comes to GenAI.

Success in this space requires a thoughtful, focused approach. M&E leaders are learning how to define GenAI’s role in their organizations, balancing immediate value with long-term industry relevance. Trusted third-party ecosystems can play a key role, helping to guide implementation, enhance business performance and mitigate risk.

The Last Word

Burger King Harnesses Fan Creativity For An AI-Powered Experience