AI for media buying: what works on Meta and Google Ads
By Matheus Mello, founder of Ads Editor and owner of YEP Agência · Published · Updated · 7 min read
AI for media buying works today in three places: reading an account in plain language with its real data, executing tasks from Claude or ChatGPT through MCP, and automating budget changes within hard limits. Strategy, offer and creative are still the media buyer's call. The useful question is not whether to use AI, but where to draw that line on a client account.
Quick answer
AI for media buying works today in three places: reading an account in plain language, executing tasks from Claude or ChatGPT through MCP, and automating budgets within hard limits. In Ads Editor's MCP logs for the 30 days to October 16, 2026, 94.6% of calls only read accounts. Strategy, offer and creative stay with the media buyer.
Summary
- AI is good at reading many numbers fast, repetitive tasks described in plain words and round-the-clock monitoring.
- It is bad at context outside the data: a sale closed in chat, stock running out, brand positioning.
- An account-aware copilot answers with the client's numbers, not with theory.
- MCP lets Claude, ChatGPT, Codex or Cursor read and change campaigns; in Ads Editor's MCP logs from September 17 to October 16, 2026, 94.6% of 96,603 calls were read-only.
- Budget automation needs a ceiling, limited steps, a stop-loss and approval for structural changes.
In this article · 7 sections
What does AI do well and badly in media buying?
| Does well | Does badly |
|---|---|
| Reading many numbers quickly and pointing out what changed | Knowing what the client sold in chat if nobody told it |
| Executing a repetitive task described in plain language | Deciding the offer and the brand's positioning |
| Watching metrics 24 hours a day | Understanding context that is not in the data, like stock running out |
| Suggesting headlines and keywords for review | Guaranteeing generated copy does not echo a competitor |
What can an account-aware AI copilot do?
The Ads Editor AI Copilot is a chat inside the manager. Unlike a generic chat, it pulls the selected account's period totals, campaigns and top ads from Meta or Google before it answers. A question such as "Why did CPA go up this week?" gets an answer grounded in that account's numbers, not in general advice.
It can also execute actions, with confirmation in the conversation, and it is built to resist malicious instructions hidden in outside content (prompt injection).
How does MCP let Claude and ChatGPT run ads?
MCP connects your AI assistant to an ad platform so you can manage campaigns in a conversation. Meta opened its own ads AI connectors in open beta on April 29, 2026, for Meta only. Google released an open-source Google Ads MCP server on October 7, 2025, and its initial release is read-only. The Ads Editor MCP covers Meta Ads and Google Ads and can act on both: it offers over 250 tools for tasks such as listing and editing campaigns, changing budgets, creating rules, reading sales by UTM, building reports and uploading media.
The setup guide, with a line-by-line comparison against Meta's official connector, is in Meta Ads MCP for Claude and ChatGPT. What people actually ask the AI to do, across 454,267 calls from 72 user accounts reported by the company, is in Meta Ads MCP usage data.
What do media buyers actually ask AI to do?
Mostly read. Ads Editor's MCP logs for September 17 to October 16, 2026, excluding the Ads Editor team's own account, show 96,603 calls from 33 user accounts across 185 different tools. The most used tools:
| Tool | What it does | Calls |
|---|---|---|
| get_campaign_metrics | Metrics of a campaign, ad set or ad for a period | 27,338 |
| get_account_dashboard | Meta account dashboard: campaigns, budgets, account status and health | 20,951 |
| list_google_campaigns | Google Ads campaigns with status and metrics | 5,151 |
| get_account_insights | Meta account totals across all campaigns for a period | 4,974 |
| update_budget | Change a campaign or ad set daily budget (most used action) | 734 |
| clone_ad | Duplicate an ad, in the same or another ad account | 596 |
| add_google_negative_keyword | Add a Google Ads negative keyword to a campaign or ad group | 484 |
Read first, act second
Of the 96,603 calls, 91,414 (94.6%) went to read tools, those whose names start with get_, list_ and similar. Actions such as changing budgets, duplicating ads, adding negatives and activating ad sets were 5.4%. Media buyers use AI to look at accounts far more than to change them.
How do you automate budgets with AI safely?
Automation without safeguards is what gives AI a bad name in ad accounts. Ads Editor AI Optimization scales and cuts budgets with a daily ceiling, steps of up to 35% per re-evaluation, pacing through the day and a stop-loss that halts scaling if results worsen on consecutive days. Structural changes, such as switching from ABO to CBO, are only suggested: the switch waits for your approval. The scaling logic is in how to scale Facebook ads.
Our position on AI touching budget
If budget automation makes you nervous, do not turn it on. The Copilot and the MCP only act when you ask. Scale by hand, in steps of up to 35%, and turn automation on for one campaign only once you trust its logic.
Before handing budgets to AI, many media buyers start with plain if-then rules they wrote themselves; templates are in Facebook Ads automated rules.
How do you start with AI without risking a client account?
- Start with read-only questions in the Copilot or over MCP: what changed this week, which ad is fatiguing.
- Move on to repetitive tasks with confirmation: pausing ads with low CTR, renaming in bulk.
- Only then test budget automation on one campaign, on the conservative profile, with a ceiling set.
Can ChatGPT or Claude run Facebook ads on their own?
They can execute what you ask through an MCP server, and some setups let them act without a human in the loop. We do not recommend that with client money, and in practice media buyers do not use it that way either. In the 30 days to October 16, 2026, 94.6% of the 96,603 calls to Ads Editor's MCP were read tools, and the most used action, changing a budget, was 734 calls, under 1%. An older, larger sample with the method behind the shares is in the usage data.
What should AI not decide?
- The offer. Price, bonus and guarantee are business decisions with consequences outside the ad account.
- The creative concept. AI can suggest headlines; knowing what the audience has seen too often comes from the market, not the dashboard.
- Structure changes on a live account. Switching budget structure resets the learning phase; a human should approve it.
- Anything that touches compliance. Health, finance and other regulated niches need a person accountable for what runs.
Frequently asked questions
Will AI replace media buyers?
No. It replaces the repetitive part: manual work, monitoring and reading numbers. Strategy, offer and creative stay with the media buyer.
How do media buyers actually use AI on ad accounts?
Mostly to read. In Ads Editor's MCP logs from September 17 to October 16, 2026, 94.6% of 96,603 calls from 33 user accounts went to read tools such as campaign metrics and account dashboards.
What is the best AI for Facebook ads?
One that sees the account's real data and acts only with confirmation. A chat without account access gives generic advice; connected through MCP or a copilot, it answers with the client's numbers.
Does the Copilot use my real account data?
Yes. It fetches totals, campaigns and top ads for the selected account, on Meta and on Google.
Can AI change my budget without me knowing?
In the Copilot and the MCP, actions are requested by you. In AI Optimization, only on the campaigns you add to it, with a ceiling, limited steps and a log of every action.
Glossary
- Copilot
- An AI chat inside the ad manager that reads the selected account's data before answering.
- MCP
- Model Context Protocol, the open standard that lets AI assistants call external tools.
- Prompt injection
- Malicious instructions hidden in data an AI reads, meant to hijack its actions.
- AI Optimization
- The Ads Editor feature that scales and cuts budgets automatically within set limits.
References
- What is the Model Context Protocol (MCP)?. Supports: MCP is an open-source standard for connecting AI applications to external systems.
- Meta ads AI connectors (Meta for Business). Supports: Announcement of April 29, 2026: connectors in open beta that let advertisers and agencies create, manage and analyze campaigns directly in the AI tools they already use.
- Open Source Google Ads API MCP Server (Google Ads Developer Blog). Supports: Released October 7, 2025; the initial release is read-only, for reporting and diagnostics, and does not make changes to the account.
The claim supported by each source is shown above.
About the author
Matheus Mello, founder of Ads Editor and owner of YEP Agência. Runs client ad accounts at YEP Agência and built Ads Editor so his own agency could stop publishing ads one at a time. Usage numbers quoted on the blog come from the product activity log. Instagram: @theusm
