Google’s 2023 Ads Safety Report: Correct Figures and Context

Google’s 2023 Ads Safety Report describes enforcement during 2023, not current-year results or a guarantee that every ad is safe. Published March 27, 2024, it reports ads, advertiser accounts, publisher pages and publisher sites as different units.

This historical summary corrects the original article’s numerical mix-up and explains how to review marketing claims without overstating what a report proves.

The corrected 2023 figures

Google’s original report gives these figures:

Swipe the table to see all columns.

These categories are not interchangeable. The earlier article’s repeated claim of 5.2 billion advertiser accounts was incorrect. Do not add the table’s values together as a count of unique scams or people protected.

What the report says about AI

Google describes machine learning and LLMs as part of its enforcement work. It also describes fraud and scam responses, election-ad verification and the Ads Transparency Center.

Those are Google’s reported activities. They do not establish an error-free system, the safety of an individual ad or the performance of BrandWell products.

What marketers should check before using a statistic

1. Original source: link to the organization that published the number.
2. Period: distinguish the measurement year from the publication and review dates.
3. Unit: label whether the number counts ads, accounts, pages, sites or people.
4. Scope: identify the platform, population and action included.
5. Comparison: use equivalent periods and definitions.
6. Claim: do not turn an enforcement count into a success rate without the appropriate denominator.

Fictional example: a review team checks 100 submitted creatives, approves 70, rejects 20 and leaves 10 unresolved. The approval rate is 70% of all submissions or about 77.8% of the 90 decided cases. Label the denominator; neither figure means 70% of all advertising on a platform is safe.

The unresolved cases are not approvals or rejections. Record the reason and next action instead of silently discarding them from a performance claim.

Review ads and landing pages together

A practical internal review can check whether the ad, destination and current offer agree:

• Can the team substantiate each factual or performance claim?
• Do the price, availability and included scope match the offer?
• Do images represent the actual product and have suitable rights?
• Does the destination work on mobile, including its form and next step?
• Are required disclosures visible and understandable?
• Is there an owner for correcting stale details?

This checklist supports review; it does not replace the advertising platform’s current policies or its approval process.

Keep evidence with the content in RankWell

For supporting articles, use RankWell to define the reader question, supply approved source material and develop the draft. Review dates, units, quotations and links before publishing through a configured destination.

Keep a claim register with the statement, source URL, date checked, scope and reviewer. When updating an old post, inspect the whole article: a correct sentence in one section does not fix a contradictory number repeated elsewhere.

Frequently asked questions

Are these current-year Google enforcement totals?

No. They are the reported 2023 figures, published in March 2024.

Does the report prove every Google ad is safe?

No. Enforcement activity is different from a guarantee about an individual ad.

Does disabling personalization remove all advertising?

Do not confuse personalization controls with removing advertising. Consult current platform instructions for the actual settings available to your account.

Reviewed and updated October 6, 2026.

Written by

Julia McCoy

Julia McCoy has contributed articles to BrandWell on content marketing, writing, and search engine optimization. This archive retains her original bylines; individual articles may be updated by the BrandWell editorial team.