Feedback Report - 2023 Q1

Quarterly report for 2023 Q1 summarizing the ecosystem feedback received on Privacy Sandbox proposals and Chrome's response.

As part of its commitments to the CMA, Google has agreed to publicly provide quarterly reports on the stakeholder engagement process for its Privacy Sandbox proposals (refer to paragraphs 12 and 17(c)(ii) of the Commitments). These Privacy Sandbox feedback summary reports are generated by aggregating feedback received by Chrome from the various sources as listed in the feedback overview, including but not limited to: GitHub Issues, the feedback form made available on privacysandbox.com, meetings with industry stakeholders, and web standards forums. Chrome welcomes the feedback received from the ecosystem and is actively exploring ways to integrate learnings into design decisions.

Feedback themes are ranked by prevalence per API. This is done by taking an aggregation of the amount of feedback that the Chrome team has received around a given theme and organizing in descending order of quantity. The common feedback themes were identified by reviewing topics of discussion from public meetings (W3C, PatCG, IETF), direct feedback, GitHub, and commonly asked questions surfacing through Google's internal teams and public forms.

More specifically, meeting minutes for web standard bodies meetings were reviewed and, for direct feedback, Google's records of 1:1 stakeholder meetings, emails received by individual engineers, the API mailing list, and the public feedback form were considered. Google then coordinated between the teams involved in these various outreach activities to determine the relative prevalence of the themes emerging in relation to each API.

The explanations of Chrome's responses to feedback were developed from published FAQs, actual responses made to issues raised by stakeholders, and determining a position specifically for the purposes of this public reporting exercise. Reflecting the current focus of development and testing, questions and feedback were received in particular with respect to Topics, FLEDGE, and Attribution Reporting APIs.

Feedback received after the end of the current reporting period may not yet have a considered Chrome response.

Glossary of acronyms

CHIPS
Cookies Having Independent Partitioned State
DSP
Demand-side Platform
FedCM
Federated Credential Management
FPS
First Party Sets
IAB
Interactive Advertising Bureau
IDP
Identity Provider
IETF
Internet Engineering Task Force
IP
Internet Protocol address
openRTB
Real-time bidding
OT
Origin Trial
PatCG
Private Advertising Technology Community Group
RP
Relying Party
SSP
Supply-side Platform
TEE
Trusted Execution Environment
UA
User Agent string
UA-CH
User-Agent Client Hints
W3C
World Wide Web Consortium
WIPB
Willful IP Blindness

General feedback, no specific API/Technology

Feedback Theme Summary Chrome Response
Testing and trialing Relevance of the testing to inform the CMA's assessment if the Privacy Sandbox APIs are not completed by the time the test begins The development of the Privacy Sandbox APIs is progressing at pace. They are already available in Origin Trial for testing and will be generally available for 100% of traffic this summer.

Additionally, we have clarified timelines for certain features (such as FLEDGE event-level reporting, FLEDGE rendering with iframes) that will not be impacted sooner than 2026.

We encourage the ecosystem to test the APIs and provide feedback to the CMA based on what testers expect to rely on once third-party cookies are deprecated. This can contribute to their assessment of the likely impact of third-party cookie deprecation.
User controls Clear guidance to ecosystem on user controls implications of Privacy Sandbox APIs We can't provide legal advice on what user controls the ecosystem can use. At the same time, Chrome is experimenting with showing updated Privacy Sandbox ("Enhanced Ad Privacy") user controls to a very small percentage of users, as part of our ongoing effort to improve the Privacy Sandbox technologies. The updates include clearer, more helpful language and layouts. Once Chrome has evaluated these refinements and decided whether to expand them to a larger population, they can share more information with the ecosystem.
Data leakage Risk of first-party data leakage to Google and other parties in the event the browser is compromised Our FLEDGE explainer makes it evident that one ad tech's data is only shared with that same ad tech (either with their worklets or their trusted servers) or when explicitly shared by that ad tech (such as a buyer shows a seller the ad URL they want to display). The one exception being that the k-anonymity check must be done by a global centralized server, which is an area we continue to devote significant resources to. Refer to the K-anonymity explainer for a detailed view of how we're thinking about privacy.

Beyond that, we are open to providing more details on how the ad-tech protections employed in the design of the k-anonymity server work.
Additional forum for discussion Request for an additional forum to the W3C for non-technical ecosystem players to share feedback The Privacy Sandbox feedback form is appropriate for general and specific comments, technical and non-technical.
The Improving Web Advertising Business Group is a forum for discussion via weekly calls and a GitHub repo.
The Privacy Sandbox Feedback page on developer.chrome.com explains other mechanisms for providing feedback and engaging in discussion. Chrome also continues to hold events such as public Office Hours to facilitate questions and content sharing. In addition, Chrome has hosted or attended more than a dozen industry events this past quarter.
Timeline clarification Clarification on the exact date for General Availability in Q3 2023 Per the timeline published on PrivacySandbox.com, General Availability is targeted to begin rollout with the release of Chrome version 115.
reCAPTCHA Impact of Sandbox APIs for reCATPCHA's spam detection use case We get feedback from reCAPTCHA periodically to ensure Privacy Sandbox proposals do not significantly impact web safety or fraud. They are developing their own plan to prepare and adjust for third-party cookie deprecation, so this question is best fielded by them.
Chrome Extensions Will Privacy Sandbox technologies such as Anti Covert Tracking (ACT) measures apply to Chrome extensions? We have not made any announcements on whether ACT may apply to Chrome extensions. However, if a technology covertly gathers information on a user, this would not align with our privacy principles.

Show Relevant Content and Ads

Topics

Feedback Theme Summary Chrome Response
TAG Design Review TAG released Early Design Review of Topics. We remain committed to Topics and have shared an update on our commitment to Topics on the latest updates page and in this issue. We responded, point-by-point, to the TAG review and shared our higher-level vision here. The Topics API will remain part of the collection of APIs that the ads ecosystem should test during 2023 — and we hope the testing feedback we hear and the implementer experience we gain will be valuable contributions in future efforts towards cross-browser standards work in this space. We look forward to continuing to engage with the ecosystem on how to ease a possible transition where Topics API could be an agreed-upon standard with cross-browser compatibility.
Approach to Topics Support for the open approach Chrome has to developing the Topics API We appreciate the sentiment and we look forward to continuing to work with the industry group to develop a Topics API that provides value to the ecosystem overall.
(Also reported in Q3 2022)
Topics taxonomy not granular enough
Broad topics taxonomy does not include more granular topics, including region specific. Q1 Update:

Improvements to the taxonomy are an ongoing effort, and in Q2 we will announce an updated taxonomy for the Topics API. To craft this new taxonomy, we worked closely with companies from across the ecosystem.
We are actively seeking feedback on the taxonomy that would be most useful for the ecosystem. In evaluating whether to expand the number of topics or include more granular topics, there are a few considerations including 1) potential privacy implications (more topics may introduce fingerprinting risk) and 2) ability to retrieve previously observed topics (such as with more topics, there may be less of a chance that an ad tech has seen the chosen topic in the past).
(Also reported in Q4 2022)
Impact on first-party signals
Topics signal may be highly valuable and as a result devalues other first-party interest-based signals. We believe interest-based advertising is an important use case for the web, and Topics is designed to support that use case. We understand that some large publishers are concerned that Topics will negatively impact their first-party data strategies. We look forward to ecosystem testing, which will provide insights into the impact Topics has on publishers.
Non-Ads related Topics use cases Use of Topics for purposes other than displaying interest-based advertising Topics is designed to address the interest-based advertising use case, which we believe is a critical use case for the free and open web. We are currently seeking feedback on other use cases and are evaluating.
Default opt-in status Regional legislation impacts for Topics consent default It is not our position to comment on legal opinions.
(Also reported in Q3 2022)
Miscategorized sites
Ads targeting when topics are miscategorized for a given site Q1 Update:
In Q2 we will announce an updated classifier for the Topics API and look forward to engaging with the ecosystem on it.
In response to the current feedback, sites are classified through a combination of a human-curated override list, containing the most popular sites, and an on-device ML model. Chrome continues to evaluate options for sites to contribute to Topics classification. Any utility improvements must be weighed against the privacy and abuse risks. For example, a few of the risks include: sites using self-labeling as a method to encode different (and potentially sensitive) meanings into topics; sites misrepresenting their topics for financial gain; sites attacking topics in order to blunt its usefulness for others (for example spamming the user's topics with meaningless noise). The public can inspect these components, with tooling available via chrome://topics-internals or this colab. Through testing, we expect classification to improve over time, and we welcome feedback of examples of sites that may be miscategorized.
Topics Classifier Request to return additional information showing the reasons when "No Topics" is returned to the caller for debugging purposes We understand and appreciate that debugging tools are helpful to developers, as they work to integrate Topics API into their systems. However, by exposing additional information (such as the reason why no Topics were returned), we may inadvertently share information that enables parties to uncover additional details (for example, if a user is in incognito mode, has disabled the API, and so forth) beyond what is intended, harming user privacy. While we don't plan to provide additional debugging tools at this time, we are open to feedback about which tools would be valuable.
Private Information Retrieval (PIR) Request for Topics API to adopt Private Information Retrieval We have previously investigated using PIR and have shared the tradeoffs here.
Bid stream Will Topics be represented distinctly from Seller Defined Audiences in the bid steam? The Topics API is a Privacy Sandbox proposal developed by Chrome, which is distinct from the IAB Tech Lab's Seller-Defined Audiences proposal. We expect the two to be represented distinctly within the bidstream. Learn how Topics will be represented in OpenRTB bid requests.

Protected Audience API (formerly FLEDGE)

Feedback Theme Summary Chrome Response
FLEDGE feature availability Clarification on timelines for testing and implementation for FLEDGE features such as Fenced Frame enforcement, K-Anonymity, and so forth. We have shared a blog post on various scoped FLEDGE features and when they will be supported. We welcome additional feedback on the announcement as we continue to develop FLEDGE.
Product rendering restrictions Request to loosen Ads Composed of Multiple Pieces restrictions for FLEDGE Fenced Frames As we announced in February, usage of Fenced Frames will remain optional until at least 2026, and iframes behavior will be supported by urn-iframes. We welcome further discussion on this topic.
Scalability issues FLEDGE performance as usage scales We are actively following up on the feedback and understanding more context so that we may propose actionable solutions. The first step was to separate the feedback into two categories, which we did:
  1. SSP-driven filtering to optimize queries-per-second (QPS) load on both a) themselves and b) the DSPs.
  2. Interest group DailyUpdate logic to optimize QPS load on DSPs.
(Also reported in Q3 2022)
Visibility of bidding logic
Concern that DSP bidding logic will be exposed in JavaScript Q1 Update:

We have shared a proposal that would limit the ability of adversaries to request data from the server in an exploratory (force browsing) fashion and we welcome ecosystem players to share their feedback or support for the proposal.
Testing difficulties Ability for smaller DSPs to properly test FLEDGE, and mitigate the risk that advertisers are only interested in testing with larger DSPs We are committed to working with smaller DSPs and strongly encourage expanded testing among DSPs and advertisers of all sizes as FLEDGE moves to general availability. We would be interested in hearing how we can best assist them in testing FLEDGE with others in the ecosystem, and welcome ideas and industry efforts to motivate advertisers to test with smaller DSPs.
Dynamic remarketing Will Dynamic remarketing still be possible with FLEDGE post third-party cookie deprecation? We are considering a response to this question and welcome ecosystem players to share additional insights on how they intend to use Dynamic remarketing.
Fraud/Abuse How can the ecosystem reduce the risks and stop bad actors or buyers from positioning themselves as a desirable audience? We look forward to engaging with ecosystem players further on fraud and abuse, and welcome more feedback in this area.
User preferences Process to save user preferences and use in ad selection For specific ads, the relevant ad tech is the party best positioned to offer controls over which creatives are shown or how they are selected.
Quantitative Testing proposal For Quantitative Testing to be fair, should the test be conducted on traffic without third-party cookies or with SSPs that only use FLEDGE? How can the mixing of signals from third-party cookies be avoided? We appreciate this feedback and are working together with the CMA to design experiments that will provide a reliable picture of the impact of third-party cookie deprecation and the introduction of the Privacy Sandbox proposals on the ecosystem. We encourage additional feedback on the CMA's Quantitative Testing proposal to be shared directly with the CMA.
Clearer documentation Request for clearer documentation on auction configuration We are hoping to share a blog post with additional overview on FLEDGE Auction Reporting in the coming weeks.
Parallelization Will the Bidding and Auction (B&A) Service support Parallelization? An ad tech using Bidding / Auction servers can start multiple servers that can serve results in parallel.
Abuse mitigation Will the FLEDGE k-anonymity server using Private State Tokens be enough to ensure user privacy? The motivation for k-anonymity is less focused on microtargeting and more on having some backstop during the interim phase where FLEDGE allows event-level reporting. We have shared more thoughts and welcome additional feedback.
ES Module conflict Request to drop generateBid as a global function as it conflicts with the ES module We are discussing this request and welcome additional feedback.
Component auction Request for publishers to have more control over auction designs Bidding and auction plan to support component auction, same as Chrome on-device.
B&A Timelines Clarity on the timeline for ad techs interested in testing B&A Servers We just updated the B&A Explainer and we updated the Timeline section to include clear definitions of timelines for different phases of Chrome-B&A testing, after aligning with the CMA.
Timeout control scheme Enhancing the timeout control scheme currently available for FLEDGE This is an interesting proposal. We will add this to the queue of proposals to study, and report on our developments.
Creative bidstreams Ability to review, and filter a winning bid, based on the creative This is an interesting proposal. We will add this to the queue of proposals to study, and report on our developments.
reportWin Proposal to provide additional information on highest scoring bid from a different interest group owner other than the winner in the reportWin function This is an interesting proposal. We will consider adding additional signals in aggregate reporting and welcome additional feedback here.
Event types Standardizing event types across measurement APIs when integrated with FLEDGE This is an interesting proposal. We will add this to the queue of proposals to study, and report on our developments. It will require coordination with our broader efforts in this field, as it would affect other Privacy Sandbox APIs beyond FLEDGE. We welcome additional feedback here.
Long-term solutions for event-level reporting Interest in keeping certain data such as highestScoringOtherBid available even after third-party cookie deprecation As we shared in the February blog post, event-level auction win reporting will be supported until "at least 2026." We do not have further details to share at the moment, but we welcome additional feedback on why it is important to keep certain data available after third-party cookie deprecation.
Interest groups limit What is the limit to the number of interest groups that an origin can add a single browser to? Chrome allows up to 1000 interest groups per owner, and up to 1000 interest group owners. These are meant to be guardrails, not to be hit in regular operation.
Event-level signals Support for a proposal to have event-level signals for generateBid and reportWin, which could be used in machine learning training We have shared our decision for browser-designed signals and ad-tech defined signals here and welcome additional feedback.
Bidding script Include user ID in the URL to the bidding script. This will not be possible as FLEDGE has the additional requirement that the tuple of the interest group owner, bidding script URL, and rendered creative must be k-anonymous for an ad to be shown.
K-anon enforcement Is k-anonymity enforced on (componentAd, size) pair? Yes, it will be. Refer to turtledove/issues/312.
Bidding and Auction Services requirements How do B&A Services support participants integrating with on-device FLEDGE and others with B&A services? We are still finalizing the design and welcome additional feedback here.
Post-view attribution Will Post-view attribution be supported? Currently we don't have any kind of standard definition of viewability and rely on the creative itself to mark a view event. Refer to turtledove/issues/452.
Lookalike targeting Can the Privacy Sandbox support "lookalike targeting?" We are discussing the use case here and welcome additional inputs.
Real-time monitoring API Proposal for a real-time FLEDGE monitoring approach We are discussing the proposal and welcome additional inputs here.
FLEDGE reporting reportWin and reportResult should be made in random order to prevent over- or under-reporting. reportResult() needs to be executed first by the seller before reportWin() by the buyer so that seller signals from reportResult() can be included in reportWin(). Refer to the explainer for more information.
Custom Key Value (K/V) servers Will custom K/V servers be supported in the future? We are discussing the question here and welcome any additional input.
Top-level auction Does one have to be the ad server to run top-level auction mechanics? The FLEDGE API does not specify which party must call it; there are no requirements in that sense in the design of FLEDGE. Anyone can run the FLEDGE auction (including multi-seller auctions). As mentioned in the Q4 2022 report, FLEDGE allows each publisher to choose the structure of the auction, including the choice of top-level and component sellers.
API scope Does FLEDGE intend to work with first-party data? We will publish content in Q2 2023 clarifying that first-party data is indeed usable with FLEDGE to both 1) use as logic to determine interest group membership and 2) to feed as user bidding signals for use in subsequent bidding logic generation.
Cross-domain interest groups Possibility of creating cross domain interest groups Any information available at the time of adding a browser to an interest group can be used to inform that audience. When third-party cookies are phased out, the availability of cross-site data to inform interest group creation will be limited.
Client-side bidding logic Porting existing server-side bidding logic to the client side We are interested in learning more about what areas are challenging or currently lacking in the porting process, and welcome any additional feedback or insights.
K/V server values Do K/V server values need to be of type string? The value needs to be a string but they can store objects in JSON or protocol buffer and serialize them into string.
Advertiser blocklist Which signals would be appropriate to provide a buyer for an advertiser blocklist? The appropriate place is either in auctionSignals or in perBuyerSignals.
Bidding unit Support for different bidding units such as CPI and CPM We would like to learn more about why this is needed given the current design and would love to hear additional feedback.
Auction logic Does the browser or the Ad server decide the winner of an auction? All winner selection is executed inside the sandbox, and all decisions are made by the seller's code. The browser simply provides a sealed, private environment inside which buyer and seller code runs.
Permissions-Policy Will the current FLEDGE Permissions-Policy continue to be enforced after the Origin Trial has ended? For the Origin Trial, the current default allowlists of both features are temporary and will be changed. We are interested in hearing how long ad techs will need to prepare for the change before we begin to enforce it.
Signal size constraint Trusted Bidding Signals requests are coalesced across multiple interest groups with the same trustedBiddingSignalsUrl; the 2MB size limit is a constraint. The constraint exists for on-device callers to prevent overwhelming resources on the device. Callers from a B&A Server will have a more relaxed constraint.
Reporting signals Add an additional signal, script-errors, to allow for the retrieval of the number of client-side errors per interest group owner and per computeBid or reportWin / reportResult. We are considering potential privacy concerns to this proposal and welcome ecosystem players to share additional insights on why this is needed.
K-Anon window size Increase the K-Anon window size from the current 7-day limit. This is under consideration and we are currently awaiting (and welcome) additional input from the ecosystem.
Device performance How does FLEDGE handle device performance if the user is in a large number of interest groups? FLEDGE offers several timeout, prioritization, and limit options across SSPs and DSPs that give ad techs fine-grained control in situations where device performance may be one reason to limit auction participation when the device is in a large number of interest groups.
B&A Services testing Request for ecosystem players to use their own server during the testing phase in order to have more logs available for debugging B&A allows users to launch and scale the servers from approved cloud providers. To maintain user privacy, we enforce execution to be done within a trusted execution environment (TEE). We are going to release an explainer about debugging of B&A TEE soon and are developing features to support that. We are seeking additional feedback on the topic.
Regulatory requirements Will FLEDGE work with cloud providers in different countries to support compliance with local regulatory requirements? We are always open to suggestions for other cloud providers, but currently we are planning at least to support GCP and AWS when third-party cookie deprecation is enforced. Refer to this explainer for more information.

Measuring Digital Ads

Attribution Reporting (and other APIs)

Feedback Theme Summary Chrome Response
Noise impact data analysis Guidance on how to perform data analysis on the impact of Noise We have shared additional documentation regarding noise and design decisions that can be used to change the impact of noise on ad-tech data.

A more detailed guide is also available.
Null reporting Clarity on the implementation of null reports We are currently working on a proposal for implementing null reports and will have more details to share soon. Implementing null reports will allow us to reduce report delays without compromising privacy.
Noise level Adjusting the noise level based on attribution window length We welcome this proposal and are looking into adding it to the specification. We welcome additional feedback here.
Trigger data size Why is the trigger data size limited to 3 bits? The size is limited to 3 bits and 8 distinct values to ensure that the amount of cross-site/context information about a user is limited. We welcome ecosystem players to submit feedback on whether the current parameterization for event-level reporting makes sense.
Event-level reporting triggers Allowing prioritization within a deduplication key We are exploring solutions to this problem and welcome additional input.
Debugging support Clarity on debugging after third-party cookie deprecation We would like to support debugging after third-party cookie deprecation and are thinking through options. We are seeking additional feedback and ideas.
Click through conversion alternatives Request for more guidance on alternatives for click through conversions We encourage the ecosystem to use the Attribution Reporting API as a durable private measurement system for applicable conversion measurement use cases. Other alternatives exist and ad tech providers will need to decide the appropriate solution based on their desired privacy and utility needs.
Billing use cases Clarity on the extent Attribution Reporting will support conversion-based billing use cases We are working on posting publicly to clarify the scope of the Attribution Reporting API for billing. The Attribution Reporting API was not initially scoped in a way that directly supports CPA billing; it supports CPC and CPM billing, which is the billing structure the majority of ad techs use.
This is something we may support in the future if there is additional ecosystem feedback.
Use case support Use case documentation for measurement API We are working on clarifying the documentation for all Privacy Sandbox reporting surfaces.
Click quality Request to add signal to distinguish intentional and unintentional clicks on an ad We are discussing the request and welcome additional input.
Measurement solution Support for measurement solutions across multiple DSPs The Attribution Reporting API can be used by measurement providers to dedupe between multiple DSPs. Additionally, we are proposing support for a list of URLs in attributionsrc, which will make it easier for DSPs to support measurement provider Attribution Reporting API requests. We welcome any additional feedback on the proposal above.
Event-level reporting Request to have the number of days before the report is sent directly available This request can already be calculated by ad techs using the currently available information. We have not heard any other ecosystem feedback regarding this request, but we are open to feedback on it.
source_registration_time Add source_registration_time in event-level Attribution Reporting. We are considering this request and welcome additional feedback on whether the ecosystem players find it a useful feature to have.
Incognito mode Will measurement solutions be available when the user is in Incognito mode? No, measurement solutions will not be available when a user is in Incognito mode. Third-party cookies are off by default in Incognito mode.
Data clean rooms Will Measurement APIs be compatible with clean rooms? A typical data clean room is an environment where individual identifier data from different sources are uploaded into a database to run analyses based on merging that underlying data. The two measurement frameworks for Privacy Sandbox APIs are event-level reports and summary reports. Event-level reports do contain an ad-tech provided event-ID that could be used in a data cleanroom, but the conversion-side information associated will be limited and noisy. Encrypted aggregatable reports cannot be used directly in a clean room, but the summary results provided by the Aggregation Service could be used as an input to analyses you perform or as supplemental information.

Aggregation Service

Feedback Theme Summary Chrome Response
(Also reported in Q4 2022)
Reporting delays
What is the expected reporting delay? Q1 2023 Update:

Following partner feedback, we have shared proposals to decrease the delay and to mitigate the impact of delay.

Both proposals have been supported by ad techs during WICG calls.
No duplicates rule How do you handle a "delayed aggregatable report" if aggregatable reports, which have the same shared ID, were already processed? We have shared a proposal on adding extra report delay to aggregatable reports' shared info and the definition of shared ID for the Aggregation Service to partially address the impact of delay loss on aggregate API. We welcome any feedback on the proposal.
Data processing Request to enable support for multiple passes of data while respecting differential privacy, using Privacy Budget We are discussing the possibility of using a more flexible way of consuming the Privacy Budget to enable this use case and welcome additional feedback.
(Also reported in Q2 2022) Query ergonomics Enable querying aggregate of keys. Q1 2023 Update:

The feature request is still being considered but we do not have any proposals to share at this time.
Origin Trial limitations Clarify the scope of the Aggregation Service such as the "no duplicates rule" which is not currently applied in the origin trial. We are looking into updating our documentation to clarify what will be available in the origin trial vs in GA.

Private Aggregation API

Feedback Theme Summary Chrome Response
Private Aggregation Contribution Budget The L1 contribution budget is too restrictive. Each call to the Private Aggregation API is called a contribution. To protect user privacy, the number of contributions which can be collected from an individual are limited.
When you sum all aggregatable values across all aggregation keys, the sum must be less than the contribution budget.

Under the current design, we set a limit on the contributions for a particular reporting origin over the last ~24 hours (as a rolling window). That's the L1 contribution budget mentioned in the feedback. We do suggest that developers scale the values they contribute based on expected volume (that is, not just using a value of 1). So, it might make sense to use a smaller value for the more common events to avoid exhausting the budget.

We're currently seeking some feedback on the Private Aggregation API's contribution budget on both the numeric bound and the scope. We are considering moving the scope from per-origin to per-site and moving the existing bound to a ten-minute window with a larger daily bound.

Limit Covert Tracking

User-Agent Reduction/User-Agent Client-Hints

Feedback Theme Summary Chrome Response
UA-R adoption Of the top 10,000 sites in the UK, only 1% of sites using programmatic advertising are sending HTTP client hints. DSPs who have not migrated may have an impact on anti-fraud capabilities. After running an analysis on the same data set, we have found that if you account for UA-CH usage via HTML <meta> tag, and the JavaScript APIs, the number of sites using UA-CH is significantly higher than the 1% figure provided in the feedback. Based on this, and other facts including ecosystem feedback, we feel confident moving forward with the gradual rollout of Phase 6 of UA Reduction, in accordance with the published timeline, while keeping the CMA informed. We note that sites have had nearly two years of lead time to prepare for the transition and a deprecation trial is still available for sites that feel they are not ready.
Hints for additional form factors Request for UA-CH to provide additional form factors such as TV, VR We welcome this proposal and are looking into incorporating it to the design. We welcome additional feedback.
Automated testing Request to resolve UA-CH bug in headless Chrome before UAR Phase 6 is shipped The bug in question has been fixed.
UA-CH support on iOS A site that relies on granular UA info for ads use cases notes that Chrome on iOS is not supported. For non-Safari iOS browsers (including Chrome on iOS), the WebKit project will need to add support for UA-CH before they can be enabled (because they control the network stack).

IP Protection (formerly Gnatcatcher)

Feedback Theme Summary Chrome Response
(Also reported in Q4) Geolocation use cases IP Protection may prevent legitimate geolocation use cases from working in the future, such as content personalisation based on geolocation. Our response is unchanged from Q4 2022:

"We are working with stakeholders to ensure that Chrome continues to support legitimate use cases for IP addresses. We are seeking ecosystem feedback on IP Geolocation granularity."
Regulatory compliance If a region has under 1M population, the current threshold of 1M for IP protection would prevent websites from using IP addresses for regulatory compliance. We are working with stakeholders to ensure that Chrome continues to support legitimate use cases for IP addresses. We are seeking ecosystem feedback on regulatory compliance on IP Protection.
Abuse mitigation Parties can circumvent IP Protection by sharing unmasked IP addresses to others. We are conscious of the risk that the current IP Protection proposal might not technically prevent parties from sharing unmasked IP addresses with others. We are working on mitigations that will avoid this risk of abuse.

As we iterate on the proposal, we encourage more feedback and discussion. Specifically, we would like to know of any use cases where parties believe they need to share unmasked IP addresses with other parties.
Network blocking Parties can circumvent network blocking by using IP Protection Proxies. The entity performing the block will need to disable IP Protection for this scenario. We have responded to the issue and welcome additional feedback.
IP address block lists impacted by IP Protection proposal Many ad tech companies utilize a basic block list of IP addresses, such as the TAG datacenter IP list, to prevent bidding on ad inventory that is highly likely to be fraudulent (or at least not monetizable). In the event an ad tech is also a tracker and could be subject to the IP Protection proposal, that company may lose the ability to perform a basic check against ads prior to purchasing advertising inventory. We encourage more feedback and discussion on the IP Protection Proposal on potential issues and solutions. One option is applying similar such lists to IP Protection, such that we are not proxying clients originating from previously flagged IP addresses.

Strengthen cross-site privacy boundaries

First-Party Sets

Feedback Theme Summary Chrome Response
(Also reported in Q4) Domain limit Request to expand the number of associated domains Our response is unchanged from Q4 2022:

"We have clarified in WICG calls that Chrome is committed to providing a usable solution that considers users' privacy interests as well. In that vein, we would appreciate feedback from the community on specific use cases that may be impacted by the domain limit, so that the team can consider ways to address these use cases while continuing to protect user privacy."
Alternative FPS submission Proposal for alternative way to submit global lists for FPS At this time, we are preparing to ship First-party Sets (FPS) in Chrome, and have set up a centralized GitHub repository to accept set submissions. Since we hope that FPS will fill a gap with existing web platform solutions in preparation for third-party cookie deprecation, we expect to learn from them how FPS is leveraged by site authors. As the list of sets grows over time, and the ecosystem adapts to a post third-party cookie world, we can also mature the process to the point where we can consider alternative decentralized schemes such as the one proposed. With the current process, we expect to institute set lifetimes, which will allow us to evolve the intake process over time. We can revisit this idea once the submission process matures.
Repo moderation Enact community moderation of the FPS Submission repo to prevent abuse. Bad actors can easily overwhelm the process employing burner origins to propose sets, and an overwhelming volume of requests might affect operations for genuine set proposals. We are trying to make the checks as objective as possible by relying on technical validation checks. We think this is the most scalable approach to the submission process. In keeping with this goal, we will also aim to ensure the process is resilient to spam / burner submissions.
Associated subsets Will FPS be able to support third-party Vendor/SaaS flow use cases through Associated subsets? The third-party vendor / SaaS flows are not a use case that is currently considered in scope for First-party Sets. We welcome additional feedback on how cross site cookies are used for these use cases.
FPS + CHIPS integration Request for FPS + CHIPS integration in order to support use cases such as A/B testing We are discussing this use case and are also considering discussing this further in a WICG call and welcome additional input here.
GDPR Proposal for a new FPS subset to be modeled after GDPR concepts We discussed this proposal internally and weighed it against other feedback received as well as our privacy goals. We've provided an answer explaining why we will not be pursuing this proposal at this time.
Memory Expected change in browser memory size when the FPS list is incorporated There have been precedents for browsers to store these kinds of lists with minimal memory impact, such as the Disconnect Tracking Protection List. While the First-party Sets list will be copied to each Chrome client locally, we will continue to monitor the file size and are confident that we can optimize the memory footprint.

Fenced Frames API

Feedback Theme Summary Chrome Response
Fenced Frames limitations Clarity around the limitations imposed by Fenced Frames In March, we updated our explainer on Fenced Frames which provides information on its capabilities and we welcome any additional feedback.
Expand access information Request to expand access to information around neighboring frames We are looking to further understand why this is a requirement from the ecosystem, and we welcome any additional feedback.
Fenced Frames and iframes Questions regarding the feature parity between Fenced Frames and iframes All available Privacy Sandbox APIs and reports will be available for iframes and for FencedFrames in the same way.
Re-sizing Fenced Frames Restricting frame size changes affects certain use cases. We are interested in learning more on the types of use cases that are affected by the restriction and would welcome additional feedback.

Shared Storage API

Feedback Theme Summary Chrome Response
Third-party worklets Can third parties write to Shared Storage, partitioned by origin? Or call other worklets for third-party measurement? The browsing context's origin of where the code is being executed determines whose shared storage that data is written to. When third-party code is added to a page, the third-party code can be embedded as an iframe with its own browsing context, which allows the third-party code to write to its own origin. The third-party code can also be embedded as a script instead of an iframe, which does not switch the browsing context, and the third-party can write to the embedder's shared storage. Note that only the owner of that shared storage can read from that shared storage.
Deduplication Deduplication would not be possible for interactions outside the Chrome ecosystem. Shared Storage is meant to provide Chrome browser-based unique reach outputs within Chrome. We are interested in working with ad techs to understand how these outputs can be used as a part of their broader reach models. We understand that the outputs themselves might only account for a portion of interactions and are interested in working with ad techs to explore additional modeling methodologies that could be layered on top.
Conversion lookback window Request to have a lookback window for conversion rate in order to see changes in conversion over time This can be implemented by processing the various conversion paths on the client side using Shared Storage, which affords additional flexibility for advanced analytics over secure unpartitioned browser storage.
Item expiry window Request to extend the expiry window to 90 days The data retention policy was updated in November 2022, and states that each key is cleared after thirty days of last write. We welcome additional feedback to understand if the new policy will work for the ecosystem.
Creative rotation Creative rotation use cases do not reflect actual actions post-auction. We are interested in hearing from more buy-side ad-tech companies on whether the creative rotation documentation is accurate or not.

CHIPs

No feedback received this quarter.

FedCM

Feedback Theme Summary Chrome Response
Identity assertion endpoint Explicitly allow arbitrary requests to the identity assertion endpoint. We have been collaborating with Mozilla on this pull request to limit websites' ability to make cross-origin credentialed requests silently without causing user annoyance and will continue reviewing and addressing other feedback as well.
Pre-populate identity Can FedCM be used to pre-populate sign in forms with an identity provider from the FedCM list? The concern for this use case is that it may result in the leaking of information when a site that has not engaged with the user is able to query the last IDP used by the user. We are discussing this issue further and welcome additional feedback.
Contextual account selection Proposal to add contextual signals in the account selection UI We are considering this proposal and welcome additional discussions.

Fight spam and fraud

Private State Token API (and other APIs)

Feedback Theme Summary Chrome Response
Capabilities gathering survey In early Q1, we finished collecting our survey results for which capabilities are needed for various anti-fraud use cases, and shared them publicly (minutes, results) We plan to incorporate this feedback as we develop new proposals and prototypes for purpose-built, privacy-preserving APIs for anti-fraud capabilities. We expect we will prioritize development where there is sufficient need and there is existing technology we can build upon to introduce the capability to the web, while preserving user privacy. For example, device and boot integrity was highly ranked and many platforms have existing APIs that securely share an assessment of device integrity, so it is a good candidate to pursue exploration within the community groups.
PST Intent to Ship feedback As part of the intent to ship, we received a concern in proceeding given that we are utilizing an older version of Privacy Pass. We also received feedback that the specification was unclear in certain sections, and should be improved to facilitate browser compatibility. We plan to implement many of the suggested spec changes before shipping to GA, as well as a few API changes. The feedback came right at the end of Q1, so we are following up on the github issues with specific details and an update to our launch plan (in progress, as of publication of this feedback report).

For the larger changes to the API, we are open to considering them, but we feel the best way forward is to proceed with launch to general availability and get hands-on feedback from more developers. We hope to continue this discussion and pursue browser standardization. If and when a new standard would emerge, we will consider adopting and developing a plan to carefully transition to it.