cd /news/ai-agents/ai-gateway-for-advertising-governed-… · home › topics › ai-agents › article
[ARTICLE · art-145474] src=snowflake.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

AI Gateway for Advertising: Governed AI Agents | Snowflake

Snowflake introduced the AI Gateway for Advertising, currently in preview, with governed Model Context Protocol integrations for Meta and TikTok and PMG as an early customer. Built on Snowflake's May 2026 acquisition of Natoma's enterprise MCP platform, the gateway connects to Cortex Agents to combine enterprise context such as business definitions and historical performance with live platform signals including delivery, catalog health and diagnostics, letting agents recommend optimizations that campaign managers approve and observe. Snowflake said additional ad platforms are coming soon.

by read5 min views2 publishedOct 5, 2026
AI Gateway for Advertising: Governed AI Agents | Snowflake
Image: Snowflake (auto-discovered)

Across marketing, AI is moving away from being just a tool that assists and is instead becoming an agent that acts, helping teams diagnose performance, decide what to do next and run while campaigns are live.

Advertising is one of the sharpest proving grounds for that shift. Yet campaign teams still move between ad platforms, pacing sheets, Slack threads and raw data to understand performance and reconcile conflicting metrics. The problem isn’t just fragmented workflows; it’s fragmented context. An agent may see a platform’s metrics without the business definitions or channel signals it needs to know what matters, leading to a recommendation that misses a core KPI or doesn’t fit the channel.

Snowflake is introducing the AI Gateway for Advertising, currently in preview, with governed Model Context Protocol (MCP) integrations for Meta and TikTok, with PMG as an early customer. Connected to Cortex Agents, the AI Gateway brings together enterprise context, including business definitions and historical performance, with live platform signals such as delivery, catalog health and diagnostics. Agents can use the AI Gateway to reason across the full picture, recommend the next best optimization and implement under policy, with campaign managers approving and observing each step.

Decide on context, act through the platform, and govern the action

The AI Gateway for Advertising, built on Snowflake’s May 2026 acquisition of Natoma’s enterprise MCP platform, enables MCP integrations with Meta and TikTok. These connections, with additional platforms coming soon, supercharge Cortex Agents to run the full campaign loop: The agent reasons over cross-channel performance, diagnoses live status and settings, decides the best action, explains the rationale to a campaign manager in the loop and acts with permission.

An agent can combine live platform signals, such as delivery, catalog health and conversion diagnostics, with data that lives in the advertiser’s Snowflake environment, such as transaction history, inventory, and the business’s definitions of return on ad spend (ROAS) and conversion. The ad platform shows what is happening in the channel, and connecting it back to data and insights in Snowflake can help articulate whether it matters to the business. Together, they give the agent the evidence and context to explain a performance change and recommend the next action, all on a shared set of definitions and KPIs.

The AI Gateway for Advertising in practice

Retail example

Let’s say a retail marketer sees ROAS drop on a live seasonal campaign. Spend is pacing correctly, and click-through rate (CTR) looks healthy. With no obvious flags, finding the cause for the drop becomes a multiday investigation that involves the ad platform, data team, catalog warnings and business data in Snowflake.

With the AI Gateway for Advertising, the marketer simply asks why. The agent combines campaign performance from the ad platform MCP with transactions, inventory and conversion-signal health in Snowflake to identify several contributing factors:

  • Purchase event quality declined after a checkout payload change
  • Several promoted SKUs have warnings
  • A high-spend product group has limited inventory

It recommends a bounded action: Reduce the affected prospecting ad set’s budget by 20%, but don’t it, as the converting cohort has strong lifetime value (LTV).

The recommendation, supporting evidence and approval are captured in one shared record, so the next person reviewing the campaign can pick up where the investigation left off.

Media example

In this example, a media marketer sees a strong response to an email promoting a new series, with one audience segment driving unusually high viewing and subscription activity. Turning that owned-media win into a paid-media action requires connecting email engagement, viewing behavior, subscription data, audience overlap and live campaign performance.

The marketer asks: Where should we increase paid-media exposure to build on this momentum? The AI Gateway for Advertising connects owned-media signals in Snowflake with live campaign data through the ad platform MCP and identifies that:

  • Viewers who clicked the email and watched the first episode are converting to subscriptions at a higher rate
  • This segment is underrepresented in the current paid audience
  • Its level is below the one associated with the strongest subscription lift
  • It would increase reach but dilute efficiency

It recommends a bounded action: Increase ad frequency for the email-engaged, Episode 1 viewer segment by 50% for the next seven days. Keep broader prospecting unchanged while the segment’s subscription and retention performance continues to validate the strategy.

The audience signal, rationale and approval are captured in one shared record, creating a repeatable handoff from owned-media insight to paid-media action.

What customers and partners are saying

Our job is turning data into decisions, and the bottleneck has never been the data — it's the context around it. Building semantic views in Snowflake Cortex AI made that context reusable. It provides one well-defined model of the business, spanning billion-row event-level data sets joined and synthesized on the fly, that any team can query in natural language and any agent can reason against. That ultimately means it takes us minutes to get from question to insight, where it used to take days of cross-functional work just to organize and query the data. The AI Gateway means those agents can now act on that insight inside our Alli platform while the campaign is live, so what used to be one analyst's finding becomes something every team can run.

Mike Treon

Head of CTV Strategy, PMG

Agentic AI is transforming marketing and it requires an interconnected ecosystem for marketers to realize its full potential. The TikTok MCP Server enables developers and platforms to build AI solutions that connect seamlessly with TikTok, helping reduce complexity, move faster, and drive real business outcomes.

David Kaufman

Global Head of Monetization Product Partnerships, TikTok

Join our session “What AI Agents Need Before They Touch a Media Decision” at Advertising Week New York on Wednesday, October 7, to see the AI Gateway for Advertising on the Tech Stage, sponsored by Snowflake.

── more in #ai-agents 4 stories · sorted by recency
── more on @snowflake 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/ai-gateway-for-adver…] indexed:0 read:5min 2026-10-05 · —