The Great Handover: Ad Platforms Are Taking Away Manual Control

For years, “AI-powered advertising” meant a feature you could opt into. In September 2026, it started meaning a decision made for you. On September 1, Google began automatically upgrading Search campaigns that use campaign-level broad match or Automatically Created Assets into AI Max, with no way to revert once the switch happens. Weeks earlier, on August 25, advertisers noticed something stranger: the placement-exclusion controls inside Meta Ads Manager had simply vanished, replaced by a bid-adjustment tool that caps how far you can discount an unwanted placement — but can no longer block it outright. Neither company made a headline announcement. Both changes were discovered by advertisers, reported by trade press, and pieced together after the fact. That, more than any single product update, is the story of marketing in late 2026: the platforms are removing the levers, and marketers are finding out in real time.

A September Full of Quiet Deadlines

Google’s AI Max rollout has been telegraphed since it exited beta earlier in 2026, but the mechanics of the migration are what make it notable. According to Marketing Dive, Google is automatically converting Dynamic Search Ads, standalone Automatically Created Assets and campaign-level broad match settings into AI Max, and is citing an average 7% lift in conversions or conversion value when the full AI Max feature set is applied versus legacy search-term matching. Dynamic Search Ads themselves — a format that has existed since 2011 — aren’t disappearing until February 2027, but the direction is unambiguous: landing-page-based targeting is being replaced by an intent-based model that blends asset signals, real-time query understanding and Google’s own judgment about what to show.

Meta’s move landed with even less warning. As reported by PPC Land, advertisers began flagging on August 25 that placement-exclusion options had disappeared from Ads Manager, with Meta’s own Help Center documentation still describing the now-missing feature as available — a gap between what the interface does and what the manual says that agencies have come to treat as a signature of how these rollouts happen. In place of exclusions, advertisers now get “value rules” that can only discount a placement’s bid by up to 90%, never remove it from delivery entirely. Meta has framed the change around performance, telling advertisers that ad sets using Advantage+ placements deliver an 11.7% lower cost per action on average than manually restricted placements.

Neither change happened in isolation. PPC Land’s reporting traces a multi-year pattern at Meta specifically: detailed targeting exclusions were removed in January 2025, a default 5% budget allocation to excluded placements was introduced that October, and the unified Advantage+ campaign structure launched in February 2026 already prohibited campaign-level placement exclusions outright. August’s change was simply the next, and most visible, step in a sequence that had been running for eighteen months.

What’s Actually Being Taken Away

It’s worth being precise about what “losing control” means here, because it isn’t really about losing access to advertising — it’s about losing the ability to say no. Under the old systems, a marketer could exclude a specific app, a specific placement, a specific device type, and know that budget simply would not flow there. Under AI Max and the new Advantage+ structure, that certainty is gone. You can make a placement expensive. You can no longer make it impossible. The same logic applies to Google’s shift away from keyword-level triggers in broad match: campaign-level broad match let advertisers control which keywords could activate expanded matching, while AI Max replaces that with an intent-based model that decides relevance on its own terms, campaign by campaign, query by query.

This is the pattern industry observers have started describing more broadly. Writing about the wider shift across social, search and commerce, marketing analyst site The Off-Label argues that brands “no longer control the middle of their own marketing” — that platforms, algorithms, creators and commerce networks now interpret and redistribute a brand’s message after the brand has only written “the first draft.” The examples it cites go beyond ad platforms: Chipotle handing an entire campaign’s creative direction to 100 independent creators with no scripts, and Zillow making housing-intent data directly available inside Pinterest’s targeting stack rather than keeping it behind its own walls. The common thread is the same one running through AI Max and Meta’s placement changes — decision-making authority is migrating from the brand’s marketing team to systems the team doesn’t fully see inside.

The Numbers Behind the Frustration

The scale of manual work that automation is supposedly replacing is itself well documented. A DoubleVerify global study of nearly 2,000 marketing and advertising decision-makers, covered by PPC Land, found that marketers spend about 26% of their working time — roughly 10 hours and 12 minutes a week — on repetitive optimization tasks like bid adjustments, budget reallocation and viewability tuning. In North America alone, that adds up to an estimated $17,000 per employee per year in lost productivity at agencies. It’s not hard to see why platforms are pitching automation as relief rather than a power grab: 42% of marketers in that same survey already use third-party AI or automated bidding tools outside their core DSPs, and another 49% plan to. The appetite for less manual toil is real. The question the industry is now wrestling with is whether the current wave of automation is actually delivering that relief, or just moving the toil into new places — like arguing with a support team about a control that quietly disappeared.

Why Marketers Don’t Fully Trust It

Adoption numbers alone tell only half the story. According to Digiday, a meaningful slice of the industry is actively pulling back from the automated products it once embraced. TJ Kropp, head of search at Ramp97, told Digiday that “when campaign performance using their AI black box dips, there is no ‘why,'” calling the lack of transparency to reproduce or improve results a significant deterrent. Mary Ann Pruitt, president and CEO of Mosaic Media, put it more bluntly: “There’s a lack of transparency and questions over where our ads are showing up… there’s a baseline of distrust.” These aren’t fringe complaints — Performance Max had already been adopted by over 95% of retail advertisers for shopping ads by late 2024, per Tinuiti data cited in the same report, and Advantage+ had crossed a $20 billion annual run rate, up 70% year over year. The scale of adoption is exactly why the trust problem matters: these products aren’t niche experiments anymore, they’re the default, and marketers are discovering the trade-offs only after they’re fully dependent on them.

That distrust is starting to show up in budget decisions. John Davis, director of audience development at Crowd Louder Media, told Digiday that some of his clients “have transitioned out of Performance Max completely, to a mix of Google ads, search campaigns and outside of the walled garden, open web campaigns.” Sara Kerr, associate media director at ZGM Modern Marketing Partners, described cutting a client’s Performance Max spend in half after finding its CPM was barely different from a manually run search campaign — automation without a corresponding performance edge, in other words, isn’t automatically worth the trade-off in visibility.

It’s Bigger Than Ad Platforms

The loss of granular control isn’t confined to paid placements. As automation absorbs targeting and creative decisions, platforms are also sharing less granular performance data back — meaning marketers lose visibility from both directions at once, per an analysis from CMSWire. And it isn’t only paid media: the same dynamic is playing out in organic discovery, where AI Overviews, AI Mode and chatbot answers increasingly decide whether a brand gets surfaced to a shopper at all, independent of how well its website ranks in a traditional results page. A brand can no longer assume that doing SEO well guarantees visibility, because the summarization layer sitting on top of search results is itself a new, largely opaque gatekeeper — the same structural shift showing up in ad delivery is showing up in organic discovery.

What Marketing Teams Should Actually Do

None of this is an argument for opting out of automation — that ship has largely sailed, and in many product categories (Search broad match, Shopping ads, Meta’s core ad sets) it is no longer optional anyway. But a few practical adjustments are worth making now:

• Audit before the deadline, not after. If your account still runs campaign-level broad match or standalone Automatically Created Assets, Google will convert it automatically within the September window regardless of whether you’ve reviewed the implications; disabling those settings beforehand is the only way to control the timing yourself.
• Stop treating platform-reported KPIs as ground truth. With granular reporting shrinking on both Google and Meta, build or strengthen an independent measurement layer — ideally anchored in first-party conversion or sales data — so a platform’s internal metrics aren’t the only signal you have when performance shifts.
• Test where manual control still exists. Retail media networks, still largely retailer-operated and increasingly offering off-site inventory, are one place advertisers retain more granular targeting control than they now have inside Meta or Google’s automated products — part of why off-site retail media spend has been growing roughly twice as fast as on-site spend.
• Invest in channels you actually own. Email, communities, events and direct-to-customer content don’t get rewritten by a platform’s automation layer, which is exactly why CMSWire’s guidance leans so heavily on building non-paid, owned touchpoints as a hedge.
• Extend the same scrutiny to AI search visibility. If AI Overviews and chatbot answers are becoming a new discovery layer, treat being cited accurately there as its own workstream, not an afterthought bolted onto existing SEO.

The Road Ahead

Google’s own migration calendar makes clear this isn’t a one-time adjustment. Dynamic Search Ads carry a separate, later deadline of February 2027, meaning marketing teams will be managing a rolling series of forced transitions rather than a single cutover — and Meta’s pattern over the past eighteen months suggests placement controls will keep narrowing in stages rather than reversing. The realistic posture for marketing leaders isn’t to wait this out; it’s to assume more manual levers will disappear on both platforms before this rolling migration is finished, and to build measurement and channel diversification that don’t depend on any single platform’s willingness to keep sharing control. The teams that adapt fastest won’t be the ones that resist automation hardest — they’ll be the ones that stopped assuming the platform’s dashboard was the whole picture months before everyone else did.