Flash Note: Meta announces ad optimization based on generative AI and user interactions
In Japan, focus may turn to Septeni (4293) …
Advertising efficiency is likely to improve dramatically through the combination of AI and user interactions.
Meta announced that it will personalize both content and ad recommendations based on interactions between users and generative AI. For example, if a user consults AI about travel, related travel ads will subsequently be displayed. At a time when an increasing number of people use generative AI as a decision-making tool, the impact of employing AI for ad optimization is expected to be substantial.
There are few empirical cases that directly measure the effect of using user prompts as advertising signals, making quantitative forecasts difficult. However, from past research we can infer the scale of potential impact. A widely cited paper, Privacy Regulation and Online Advertising (Management Science, 2011), used large-scale empirical data to show that after the EU’s “Privacy and Electronic Communications Directive” was enacted, the effectiveness of online display advertising in terms of “change in purchase intent” fell by an average of 65% compared to countries without such regulation. If regulation produced such a significant negative impact, then conversely, leveraging user prompt data—so deeply tied to behavior—could yield extremely positive effects.
Implications for the media and advertising industry
According to Dentsu, Japan’s total advertising expenditure in 2024 was JPY 7.673 trillion. Of this, internet advertising accounted for JPY 3.6517 trillion, or about 47.6% of the total. In 2014, this ratio was only 17.1%, underscoring the explosive growth of digital. A similar shift is now underway, and in the medium term AI agents may erode market share not only from television but also from conventional internet platforms.
For corporate advertisers and agencies with high AI literacy, however, this transition could be an opportunity. In Japan, Septeni (4293) has been notable for its advanced collaboration with Meta. (A majority of Septeni’s shares are held by Dentsu, 4324.) While this is framed more as content optimization rather than direct prompt utilization, Septeni’s achievements with Meta’s ASC (Advantage+ Shopping Campaigns) earned it the title of Best AI Solution Partner from Meta in 2024. It is therefore reasonable to expect that Septeni will play a significant role within Meta’s broader ad optimization strategy.
Background: Why AI platforms are accelerating monetization
Turning back to the U.S., around the same time as Meta’s announcement, ChatGPT revealed plans to launch a new TikTok-like service, highlighting the acceleration of “downstream monetization” strategies by generative AI platforms. Several drivers explain this push.
First, the surge in computational costs is severe. Developing, training, and running large language models (LLMs) requires enormous computing resources. Investment in this area is also rising; for example, in September 2025 Nvidia announced it would invest up to USD 100 billion (approx. JPY 15 trillion) into OpenAI.
Second, rising legal and ethical risks are increasing compliance costs. The sourcing of massive datasets for model training has already sparked lawsuits over personal data, privacy, data scraping, and intellectual property infringement, with more expected to come. New regulatory frameworks such as the EU AI Act (which classifies AI systems by risk and imposes restrictions) and the NIST AI RMF (a U.S. government framework for AI use) will require further compliance investment.
Against this backdrop of escalating costs, AI platforms are rushing to monetize, and as a result, the dynamics within the advertising industry are expected to accelerate further.