Google’s second September spam-update wave, AI Mode monitoring and checkout, Gemini 4 Argon, OpenAI Dots, and the effect of AI search on retail and publishers.

Google’s September spam update, multimodal Search Console reporting, the AI Mode click experiment, and new GPT-6 and Claude models.

This article is also available in Chinese: 中文版.

Many teams working on cross-border marketing and GEO (generative engine optimization) are asking the same question. ChatGPT, Perplexity and Google’s AI answers cite Reddit all the time, so will a brand show up in those answers if it posts, comments and upvotes more on Reddit?

On July 6, 2026, Reddit published a set of anti-spam numbers. It now blocks about 23 million spam views per day before the spam reaches a person, catches about 25,000 new spammy posts and comments per day, and has revoked nearly 2 million inauthentic votes per day over the last three months. As a result, mass posting and fake user discussions are less and less likely to work on Reddit. Brands should first study the real discussions that already exist on Reddit, and then take part under a transparent identity.