Brands need an AI search playbook, or they’ll put their reputations at risk

As AI search continues its rapid ascent, it presents a clear and present risk to brand reputation and corporate sales. Yet, many companies remain unaware of what large language models (LLMs) like Open AI’s ChatGPT, Google Gemini and Amazon’s Rufus are saying about them.

Do you know your brand’s visibility? Favorability? Share of voice? How do you compare to your competitors? Are there differences in the responses between Rufus and OpenAI? Spoiler: Answers vary significantly between AI models.

One study last year found that 86% of Google searches return results with generative elements, and Gartner forecasts a 25% decline in traditional search volume as AI search rises. This disruption is no longer theoretical — it’s happening. And fast.

To truly grasp the scale of this threat, brands need to understand the shift in how information is retrieved by the LLMs. Traditional search tools like Google Search analyze content and websites and return a ranked list of clickable links based on authority.

In contrast, AI search functions quite differently. Here, an LLM analyzes the sites and content to understand the context of a query, then crafts a response to the user’s query. In this new paradigm, it is no longer about ranking No. 1 in search, but rather ensuring your product or service is embedded in the LLM’s answer.

This introduces risk. LLMs are trained on massive datasets that include user-generated content, news, forums and the open web, and the majority are unverified. Even the most advanced LLM models hallucinate — or make up answers. According to MIT Technology Review, OpenAI’s latest GPT-4.5 model made up answers 37% of the time on the SimpleQA benchmark test. GPT-4o and o3-mini fared worse, with hallucination rates of 59.8% and 80.3%, respectively.

As AI search becomes more mainstream, users will likely accept AI-generated answers at face value. This compounds the reputational risk and introduces the possibility of product misuse based on incorrect information.

LLMs are evolving and prioritizing third-party, highly trusted and contextualized content over keyword-stuffed pages to counter this. They are also crafting deals with premium publishers to ensure access to high-quality journalism. Net, AI search is increasingly tied to narrative authority, not keyword density.

The shift from clicks to narratives is seismic. Search is transforming from a performance-driven funnel to a brand-building channel. Agentic AI, which are autonomous systems like OpenAI’s Operator, will likely accelerate this further. These changes necessitate the need to actively understand and manage a company’s AI presence. Are you visible? Are your narratives favorable? Is the information about you accurate? If not, your competitors might be.

This calls for a new playbook. Traditional SEO won’t work. Instead, brands need to prioritize having communications experts with a deep expertise in narrative construction, authority and trust building, reputation management and influence at scale. While technical SEO, content and performance marketing are still critical, their importance will diminish as Generative Engine Optimization (GEO) — the process to optimize AI search — takes precedence.

If you own or are responsible for a company, product or service, here are three things you should do today to help prepare:

• Conduct an AI audit: Don’t rely on a cursory check. Make sure it is a robust analysis of at least three platforms and tens of thousands of queries for a minimum of a few weeks. This will allow you the depth of size and scale to truly understand your visibility, narrative sentiment and comparative standing.

• Identify gaps and optimize: If your product or service isn’t visible, what is? What narratives dominate? Use these insights to restructure your brand.com, product detail pages (PDPs) and FAQ content to align with what LLMs prioritize — clarity, context and authority.

• Synchronize paid, owned and earned media: Ensure all channels work cohesively to influence GEO outcomes. Messaging consistency, narrative control and authority sourcing are now intertwined with performance.

SEO may still dominate today, but not for long. Inaction is not a strategy. Brands that fail to adapt to AI search risk being misrepresented, overlooked or replaced entirely in the AI- generated answers their audiences now trust.

This op-ed represents the views and opinions of the author and not of The Current, a division of The Trade Desk, or The Trade Desk. The appearance of the op-ed on The Current does not constitute an endorsement by The Current or The Trade Desk.