Google Ads Advisor Bug: MCC Chats May Mix Client Context

Google Investigating Google’s new AI-powered Ads Advisor tool is under scrutiny after advertisers noticed that chat responses inside one client account may draw context from conversations held in entirely different client accounts within the same MCC. For agencies managing dozens or even hundreds of advertisers, the discovery raises immediate questions about data separation, client confidentiality, and the reliability of AI-generated optimization advice. The issue was first raised by PPC expert David Melamed, who shared that Ads Advisor seemed to “carry over” strategic context from one advertiser to another when the same user handled both accounts under an MCC. According to him, the chatbot referenced business details from one client when he opened a fresh chat inside a different client’s account. Although the examples contained confidential data and were not publicly shared, Melamed confirmed that the overlap was “very noticeable.” Google Ads Liaison Ginny Marvin acknowledged the reports, saying, “Thanks for flagging. The team is looking into it.” It strongly suggests that the behaviour is unintentional — likely a memory, session, or scoping issue within the AI assistant’s architecture. Key Highlights Google Ads Advisor may mix chat context across MCC accounts, causing cross-client data and strategy overlap for agencies. Google has acknowledged the issue, with Ads Liaison Ginny Marvin confirming that the team is investigating the unintended behaviour. Advertisers are advised to avoid sharing sensitive information with Ads Advisor until Google issues a fix for the context-bleed bug. Why This Matters for Agencies and Advertisers For Google Ads users with MCC structures, this bug has two critical implications: 1. Cross-Account Data Leakage Agency strategists often discuss budgets, conversion issues, competitor insights, and campaign structures inside Ads Advisor. If that information can appear in unrelated accounts, confidential data may unintentionally surface. 2. Unreliable Optimization Advice Suppose another advertiser’s industry or goals influence the AI’s recommendations. In that case, users may receive irrelevant or misleading guidance, which can be hazardous for accounts in regulated sectors such as healthcare, finance, or legal services. Related Services: Google Ads Services Why It’s Likely Happening While Google has not published technical details, early signs indicate that Ads Advisor may be: Using MCC-level conversation history instead of isolating context per account caching user-level chat sessions sharing embeddings across advertisers under the same manager login Given that Google only launched Ads Advisor weeks ago, the behaviour is likely due to early-stage bugs in integrating Gemini-powered AI across the Ads UI. What Advertisers Should Do Now Until Google issues an official fix: Avoid sharing sensitive client details with Ads Advisor Treat all chats as if they may influence other accounts. Use external documents or internal notes for strategic discussions. Report unusual behaviour directly to Google support Bottom Line Google’s investigation is ongoing, but advertisers — especially agencies — should remain cautious. Until the patch is confirmed, Ads Advisor may unintentionally mix context from one MCC sub-account into another, potentially affecting both privacy and optimization guidance. Also Read: Google Ads MCC Takeover Attacks Surge

Google Ads Advisor Bug May Mix Context Across MCC Client Chats

Continue Read Other Related News