On July 1, 2026, Google updated its Google Ads Terms of Service for the first time in eight years. The update was announced in advance, requires no action to accept, and applies automatically to every advertiser using the platform. Most accounts received a notification in June; the new terms are now in effect.
For B2B SaaS advertisers, three specific changes matter more than the rest. They concern how Google uses the data you enter into AI-powered tools, what domain crawl authorization actually grants, and where liability sits when AI-generated assets create compliance problems. None of these are hypothetical — all three reflect behaviors Google's automated systems already exhibit, and the updated terms clarify the legal framework around them. This post works through each one and what it means for a typical SaaS Google Ads account.
AI tool inputs and the shared learning system
The most significant language change covers what happens to data you enter into Google Ads' conversational and AI-powered tools — the AI Max setup flows, Gemini-powered asset generation, and similar experiences that have become part of the standard campaign creation interface. As ALM Corp summarized the update, "information entered into conversational experiences and similar Google Ads tools may be utilized by Google's systems" — with the key addition that this use extends to improving how those systems work for all advertisers, not just your account.
The prior terms were narrower: inputs to your campaigns improved your campaigns. The updated terms reflect the reality of how large AI systems actually work — inputs from millions of advertisers collectively train and refine the models that everyone uses. Google does not sell or expose one advertiser's data to another. But when you type your ideal customer profile into an AI Max setup flow, describe your target audience to a Gemini-powered asset generator, or enter competitive positioning into an automated briefing tool, that input can inform the underlying systems that determine how Google's AI interprets and responds to similar inputs from any advertiser. For most B2B SaaS companies, this is an acceptable trade-off. The caution is at the margins: avoid entering genuinely proprietary competitive intelligence, unpublished pricing structures, or strategic positioning that you would not want incorporated into a shared model.
Practically, this means treating AI setup inputs as you would any third-party tool — enter what is necessary for the task, be deliberate about what is sensitive, and prefer publicly available language from your site over bespoke competitive framing typed ad hoc.
Domain crawl authorization: what you are actually granting
The second material change clarifies what happens when you authorize Google to crawl your domain — a permission that is prompted during AI Max setup and some Performance Max onboarding flows, often as a frictionless checkbox. As the updated terms now make explicit and ppc.land noted: "Google can use a landing page URL to generate campaign structure, inspect destination content, and produce related assets from what it finds there. If you authorize Google to crawl a domain, you are granting access to public content that may be recast into campaign output."
For a well-maintained SaaS website, this is often beneficial — your homepage, feature pages, and case studies contain exactly the messaging that should be in your ads. The risk is in the edges of a typical SaaS site: outdated pricing pages that have not been retired from the sitemap, blog posts describing features that have since changed, beta feature pages visible to crawlers but not meant for ad copy, or landing page variants that contain experimental claims. Any of these can become AI- generated asset inputs once crawl authorization is active.
The remediation is straightforward. First, audit your account settings to confirm whether domain crawl is enabled and for which domains. Second, review your site for content you would not want converted into ad copy — outdated pricing, deprecated features, beta language — and either retire those pages or add them to a disallow rule in your robots.txt. Third, build a habit of reviewing asset groups after any automated asset generation, since the crawl-based output may not match your current messaging. The good news is that asset-level controls exist — you can remove individual AI-generated assets that do not reflect current positioning without disabling automation entirely. For a structured approach to what to check, our Performance Max audit guide covers asset group review as a core step.
Liability for AI-generated content: what it means operationally
The third change is the one with the sharpest edge for B2B SaaS teams. As ZATO PPC Marketing summarized the terms update: "The updated terms appear to give Google broader authority to use automated systems and AI to generate, select, and optimize campaign elements on advertisers' behalf. At the same time, advertisers remain responsible for reviewing and approving those campaigns and any resulting ad assets." If an AI-generated asset violates policy or makes a misleading claim, the advertiser bears the liability.
For B2B SaaS specifically, this is not an abstract risk. Common SaaS ad copy includes claims that change: integration counts ("connects with 200+ tools"), uptime guarantees ("99.9% SLA"), compliance certifications ("SOC 2 Type II"), pricing ("starts at $X/month"), and competitive positioning ("the only platform that..."). Google's AI-generated assets pull these from your site and from inputs you have provided. If the site has not been updated to reflect a changed integration count, or if a pricing page is mid-revision, an AI-generated headline can end up making a claim that is outdated the moment it runs. The fact that Google's system generated it does not move the compliance obligation.
The operational response is to treat AI-generated assets as you would any copy submitted for legal or marketing review — with the recognition that the volume is higher and the review cadence needs to match. For most SaaS accounts, a weekly asset audit covering newly added AI-generated headlines and descriptions is sufficient. For accounts with active regulatory exposure — fintech, healthcare, enterprise SaaS with strict compliance requirements — that cadence should be tighter. Enabling asset-level approvals where your campaign structure allows, and regularly cross-referencing AI-generated claims against your current public documentation, are the two most practical controls available today.
What is not changing — and what was already true
It is worth stating what the July 2026 TOS update does not do. It does not give Google permission to sell your account data to competitors. It does not enable Google to run ads on your behalf without any input from your account. It does not change the core structure of the advertiser- platform relationship — you still control campaign budgets, targeting, and the assets you explicitly create. What it does is codify, in legal language, the behaviors that AI Max, Gemini-powered tools, and Performance Max's automated asset generation already exhibit. The crawl-based asset generation, the shared learning from AI tool inputs, and the advertiser-side liability for generated content were all effectively true before July 1 — the update makes them unambiguous.
Search Engine Land noted that this is the first TOS overhaul in eight years — a span that covers the entirety of Google's shift toward automated and AI-driven ad delivery. The update reflects how advertisers' inputs are used across AI-powered and automated advertising tools, rather than introducing fundamentally new capabilities. For B2B SaaS advertisers who have been running AI Max, using Gemini for asset generation, or running broad Performance Max campaigns, the behavioral reality has not changed — only the contractual clarity around it.
SaaS-specific risks the updated terms amplify
Several characteristics of B2B SaaS businesses make the updated terms particularly relevant. SaaS products change faster than most — features are added and deprecated quarterly, pricing models evolve, certifications are obtained and maintained. A site that accurately reflected your product six months ago may contain outdated claims that, when crawled and auto-converted into ad copy, create genuine compliance exposure. The solution is not to disable automation but to maintain sharper site hygiene: ensure pricing pages reflect current packages, retire pages for deprecated features, and update compliance claims the moment a certification changes.
B2B SaaS companies also tend to have complex buyer journeys with multiple audiences — a data engineering persona, a VP of Engineering, a procurement team. AI-generated assets optimized for click-through rate may blend messaging across these segments in ways that create ambiguity about who the product is for and what it does. This is a quality issue as much as a compliance one, and it is the kind of drift that happens gradually across dozens of AI-generated asset variants. The enhanced conversions updates from earlier in 2026 already required B2B SaaS teams to be more deliberate about what they feed the bidding algorithm; the TOS update extends that same deliberateness to what they feed the creative generation layer.
Three things to do this week
The practical response to the July 2026 TOS update is a focused audit, not a platform overhaul. Three things are worth doing in the next week. First, check your domain crawl authorization setting — find it under your account settings and confirm which domains are enabled. If you have not deliberately activated it, confirm the status regardless, since onboarding flows sometimes enable it by default. Second, walk through your site as if you were a crawler looking for ad copy: identify any pages that contain claims that are outdated, experimental, or too nuanced to survive compression into a headline. Third, pull your current AI-generated assets in your active campaigns and compare each claim against your current product documentation — flag any that no longer match your current public position.
This is not a one-time task. The updated terms describe an ongoing dynamic: as long as you use AI-powered tools and domain crawl authorization, the generated output will continue to reflect whatever Google finds. Building a lightweight monthly review into your account maintenance rhythm — checking AI-generated assets the same way you would check any live creative — is the appropriate response. If your account relies heavily on AI Max and automated asset generation, the transition to treating asset review as a recurring line item is exactly the kind of governance shift the updated terms are designed to prompt. For a broader view of what automated tools are actually doing in your account, the AI Max migration guide covers the behaviors most relevant to SaaS advertisers. And if you want a structured review of how your account handles AI-generated content today, our Google Ads audit includes an asset and automation review as a standard section.