From Conversations to Cash: Defining a New AI Model
AI chatbot monetisation through contextual advertising is a business model where conversational AI systems earn revenue by inserting relevant ads directly into their AI-generated answers, matching commercial offers to the user’s intent in real time while keeping access free or low cost. Taboola’s DeeperDive answer engine sits at the centre of this shift, turning the natural flow of user questions into a steady stream of high-intent ad impressions. Instead of relying on display banners around static articles, DeeperDive builds ads into the conversational response itself, which is closer to how users now search, compare and learn. Generative AI companies, chatbots and virtual assistants can connect to this infrastructure and earn a share of the advertising revenue, tying their income to the quality and volume of user engagement rather than to subscriptions alone.
Inside Taboola’s DeeperDive Monetisation Engine
DeeperDive was originally built as an AI-powered answer engine for publishers, using their editorial archives to generate conversational responses. It now adds a new monetisation channel by inserting high-intent ads directly into the AI-powered results page, turning user inquiries into commercial opportunities. According to Taboola, DeeperDive combines large language models, retrieval systems and its proprietary intent graph to generate tens of millions of AI-powered answers every month for more than 7 million users. The system is powered by NVIDIA accelerated computing so it can deliver high-throughput, low-latency responses at scale. On the demand side, Realize, Taboola’s performance ad platform, feeds ads from tens of thousands of advertisers into these AI answers. This infrastructure lets AI answer engines match ads to context in milliseconds, without forcing publishers or AI providers to build their own ad stack.
AI-Generated Ad Placement in Action
The most visible change DeeperDive brings is AI-generated ad placement directly inside the conversation. When a user asks about buying a home, for example, the AI can respond with an explanation of mortgage options while displaying a mortgage-related advertisement beside or within the answer. This is conversational AI advertising in a literal sense: the ads appear where the user’s intent is clearest, not on a generic homepage. Because the engine is trained on editorial content and linked to Taboola’s understanding of user behaviour, it can distinguish between informational and commercial moments and show ads when they are most welcome. For AI answer engines, this model turns each question into potential answer engine revenue, keeping the focus on relevance and intent rather than on broad demographic targeting or intrusive interruptions.
Beyond Publishers: Opening the Stack to AI Platforms
Taboola is now opening DeeperDive’s monetisation engine beyond traditional publishers to generative AI companies, including conversational AI platforms, chatbots and virtual assistants. This marks a shift from a publisher-only ad network to an ecosystem where any AI agent can plug in and monetise user queries. Adam Singolda, Taboola’s CEO, describes the move as “helping build the economic layer of the AI internet,” arguing that the future will be “increasingly conversational and agentic.” Many AI companies have built products people like, but still lack reliable business models. Because users will not subscribe to every single AI service they try, advertising and commerce become natural ways to fund ongoing innovation. By sharing DeeperDive’s infrastructure, Taboola gives AI builders an off-the-shelf monetisation stack instead of forcing them to negotiate their own advertiser relationships.
What This Means for the Future of AI Business Models
Opening DeeperDive to generative AI providers signals how AI chatbot monetisation may evolve: conversational AI advertising becomes the default, not the exception. AI services can remain free or low-cost because answer engine revenue comes from relevant commercial messages that appear within AI responses. This aligns with how users already interact with AI tools, turning natural questions into funded interactions rather than paywalled content. For advertisers, DeeperDive promises access to high-intent moments at scale, where users actively seek advice or solutions. For publishers, it keeps AI answer engines tied to trusted content while adding incremental income. If this model spreads, AI-generated ad placement could become a standard feature of chatbots and virtual assistants, encouraging an internet where the most helpful answers are subsidised by ads that feel contextual rather than intrusive.
