“We have been toying with what to call it,” says David Kaganovsky, global chief technology officer at Landor, a WPP company. Kaganovsky would call it an “Intelligent Brand System.”

The Intelligent Brand System would act as a safety check to define the company and its brands. The if-then rules would allow these guidelines to work dynamically.

“We've been working with our partners at WPP to create a consistent taxonomy, a way of defining and speaking about a brand,” he said. “The tools we're in the process of developing will allow me to take a brand's guidelines and evaluate it against AI and the taxonomy we defined.”

People must have the ability to understand the rules without misinterpretation, he said. It must have a lot of dos and don'ts, and become a dataset that can inform the brand's rules.

This process requires a decisioning engine and a way to interpret the brand and feed it back into the system.

Once the guidelines are AI-ready, it would become a service and be digitally called up many times per second.

It could be embedded into media, production, or advertising workflow. As an ad is created, this system would check the guidelines, and potentially, in a dynamic media environment.

Advertisers will have the ability to automatically check the ad's message or creative design against the brand's rules, for example.

“If the rules say when the weather is this, or geopolitical situation is that, or the news says this, the technology will have the ability to make the decisions,” he said. “We need to make these decisions quickly.”

Kaganovsky called it the perfect use of AI for agencies and brands because the technology can handle massive amounts of data and can quickly interpret the information.

Years ago, these types of rules could be found in a book on someone's desk, but these days it's a PDF on a computer in as little as two pages, Kaganovsky said. Now companies have a new opportunity.  

“We are developing agents to use for a variety of tasks every day,” he said.

Landor has dozens of multiagent systems known as agentic. Kaganovsky rattled off some of the agents or agentic technologies that Zandor developed. The list includes Brand Touchpoint Intelligence, which analyzes a brand's touchpoints and scores them by comparing it to competitors much faster than humans; and Knowledge Analyzer, which allows humans to ask it questions, and the agent goes out across WPP's network to find the answers. It doesn't assume knowledge.

Matt Kissane, Landor global executive director, in a blog post explored how to prepare brand guidelines for the future of intelligent marketing. One of the most important points he made is that “assumed knowledge” is a brand's enemy.

“Humans, especially experts, tend to assume a common level of knowledge,” he wrote. “For designers and brand experts, these are unwritten principles that they absorbed at design school, or over the years at work in an agency.”

AI is a novice. It may have been trained on data, but it never went to design school or worked at a brand agency. It’s only as good as the data fed into the system to learn.

Agentic capabilities will become more widely available in everything from ad targeting to shopping.

Google announced at the Google Next conference in April that they were working on a protocol called Agent to Agent (A2A). The protocol allows agents within Salesforce's Agentforce ecosystem to interact with each other as well as external agents.

It focuses on areas such as authentication, identification and message passing.