Gemini model capabilities for Places API (New) brings support for LLMs to the Places API so that you can return the data associated with a place (opening hours, reviews, photos) along with data provided by LLMs. This generative data contains aggregated and summarized data from the LLMs to enhance the data already associated with a place.
For example, a standard Text Search (New) for "EV charging stations" returns a list of places that offer EV charging. This type of response is fine if you are just searching for the nearest EV charger or the nearest one compatible with your vehicle.
But you might also want to get a better understanding of the area around the charging station, especially if you are going to be there for an hour or more while your vehicle charges. When using the Gemini model capabilities for that same search, the response can contain dynamic, generative descriptions of each place in the response, including information about the area around the charging station:
"Within a 2-4 minute walk, there are several dining and shopping options, including Chef Chu's, Gen Korean BBQ House, Amber India, Pacific Catch, and several grocery stores like Walmart, Whole Foods Market, and Trader Joe's."
"Within a 1-5 minute walk, there are several dining and shopping options, including Khao Kang Thai Kitchen, Roger's Deli & Donuts, and La Costeña. For sushi, Joy Sushi is a good option. For groceries, Dollar Tree offers a variety of products at discount prices."
This generative content from the LLMs enhances the response and provides additional information to help users make an informed decision.
New AI-powered features added to the Places API (New)
These capabilities add the following data to a Places API (New) response:
Generative place summaries which are related to a specific place. Place summaries are supported by Place Details (New), Text Search (New), and Nearby Search (New).
Generative area summaries which are related to a specific area around a place. Area summaries are supported by Place Details (New), Text Search (New), and Nearby Search (New).
Contextual search results are supported by Text Search (New) and contain data based on the search string passed in the request, instead of data directly associated with a place. This data helps to understand how Text Search (New) determines the response.
Generative place summaries
Generative place summaries are specific to a place ID. Place summaries aggregate many different types of data into a single summary for the place.
Places API (New) provides two types of generative place summaries:
Overview summaries: Brief overview of a place, on average around 100 characters, that can help users get a snapshot of a place.
Description summaries: Longer summaries of a place, on average around 400 characters, that provide a detailed overview covering the top aspects of the place.
Place summaries are currently available for place types shown in Supported types for the categories Culture, Entertainment and Recreation, Food and Drink, Shopping, and Sports.
Generative area summaries
Generative area summaries are created by the LLM for the nearby area around a place. The area summary for a place might contain all the information necessary for a user to make a decision, without the user having to make further requests to drill down into area details.
Along with a description of the area, the response also contains a list of place IDs for the places referenced in the description, and a list of places sorted by different content types. For example, one content type contains a list of places associated with activities in the area, and another contains a list of restaurants in the area.
Area summaries are currently available for place types shown in
Supported types
for the Lodging category, as well as the specific place types
electric_vehicle_charging_station, geocode, and premise.
Place and area summaries generation
Both the generative place summaries and the generative area summaries are generated offline frequently by Google. By generating these summaries offline, you can take advantage of the content powered by Gemini while still seeing the same performance from the Places API (New) that you already expect.
Google frequently regenerates these summaries to ensure that they are fresh based on the latest available information. That means you will display the freshest data in your service when you call the API.
Contextual search results
A Text Search (New) returns information about a set of places based on a string — for example "pizza in New York" or "shoe stores near Ottawa". The Text Search (New) response contains a list of places matching the text string and any location bias set in the request.
When you add contextual search results to the response, the response now include reviews, photos, and justifications that help users answer the questions:
- Why am I seeing this place in the results?
- Why should I choose this place?
For example, you make a request to Place Details (New), Text Search (New), or Nearby Search (New) today that includes a specific place in the response. The response data for the place is the same for all three requests because the returned data is associated with that place.
However, for a Text Search (New) using the new capabilities, the response can include data derived from the input query string itself. That means the contextual search results are unique to the query string. For example, you make a Text Search (New) request with the following query string:
"Spicy vegetarian restaurant in Mountain View, CA"
You then leave all input parameters the same and only change the query string to:
"Spicy vegan restaurant in Mountain View, CA"
For a restaurant that appears in both responses, the response data about the place is the same because that data is associated with the place itself. But the contextual search results are unique to the query because they are derived from the query string. You can then display the reviews, photos, and justifications from the contextual search results to your users to help them decide which place in the response is most applicable to their search request.
Enable Gemini model capabilities for Places API (New)
You must enable Places API (New) to use Gemini model capabilities for Places API (New). For more information, see Set up your Google Cloud project.
Restrictions for the experimental release
This release only supports place and area summaries for places in the United States in English only.
What is Gemini?
To learn about the Gemini model, see the Gemini FAQ.
Request a generative place summary
Place summaries are generative summaries created by the LLM for a specific place. Places API (New) adds two types of summaries to a response:
Overview generative summaries that are similar to the static editorial summary associated with a place. Generative short summaries, on average around 100 characters, have several advantages:
Many places don't have editorial summaries, so they fill a gap in the information available for a place.
They are generated from the LLM so there is no need for the place owner to maintain them.
Description generative summaries are derived from many sources, including reviews from Google Maps. The description, on average around 400 characters, includes a list of reviews used to generate the long description.
To generate description summaries by hand requires a person to aggregate data from many locations and then hand-craft and maintain the content. Gemini model capabilities for Places API (New) build these summaries from the LLMs to maintain the scope and freshness of the generated content.
Place summaries are supported by Place Details (New), Text Search (New), and Nearby Search (New). To include the place summary in the response, include the following field in the field mask of the request:
Place Details:
generativeSummaryText Search and Nearby Search:
places.generativeSummary
The
generativeSummary
field contains the following fields:
generativeSummary.overview: The generative overview summary.generativeSummary.description: The generative description summary.generativeSummary.references: An array of up to five reviews used to provide content for the description summary. If there is no description summary, this array is empty.
Example overview summary request
This example Text Search returns just the overview summary for spicy
vegetarian restaurants in Mountain View, CA:
curl -X POST -d '{
"textQuery": "Spicy Vegetarian Food",
"location_bias":{
"rectangle":{
"low": {
"latitude": 37.415,
"longitude": -122.091},
"high":{
"latitude": 37.429,
"longitude": -122.065}
}
},
"maxResultCount": 5
}' \
-H 'Content-Type: application/json' -H "X-Goog-Api-Key: API_KEY" \
-H "X-Goog-FieldMask: places.id,places.displayName,places.generativeSummary.overview" \
https://places.googleapis.com/v1/places:searchText
The response contains the generativeSummary.overview field which includes the
overview.text and the overview.languageCode:
{
"places": [
{
"id": "ChIJQ8nPjFW3j4ARcAaUg5pF-Ts",
"displayName": {
"text": "Zareen's",
"languageCode": "en"
},
"generativeSummary": {
"overview": {
"text": "Casual restaurant serving halal Pakistani and Indian cuisine, plus a selection of vegetarian items.",
"languageCode": "en-US"
}
}
},
{
"id": "ChIJETHw0F-7j4AR85V_I6OobCI",
"displayName": {
"text": "Hyderabad Dum Biryani",
"languageCode": "en"
},
"generativeSummary": {
"overview": {
"text": "Casual eatery offering vegetarian & non-vegetarian Indian fare, plus breads & desserts.",
"languageCode": "en-US"
}
}
},
. . .
}
Example overview and description summary request
This example makes the same query as shown in the previous example but uses a
field mask of places.generativeSummary to return the entire
generativeSummary object, including the generativeSummary.overview,
generativeSummary.description, and generativeSummary.references fields:
{
"places": [
{
"id": "ChIJQ8nPjFW3j4ARcAaUg5pF-Ts",
"displayName": {
"text": "Zareen's",
"languageCode": "en"
},
"generativeSummary": {
"overview": {
"text": "Casual restaurant serving halal Pakistani and Indian cuisine, plus a selection of vegetarian items.",
"languageCode": "en-US"
},
"description": {
"text": "Pakistani restaurant dishing up comfort food including vegetarian, vegan, and halal options.
Offerings include chicken tikka masala, naan, and paratha. There's also a salad bar, and dessert is
available.\nOpen for breakfast, lunch, and dinner, the venue has a casual atmosphere and outdoor rooftop
seating. People mention the ambiance and say the service is prompt.\nCustomers typically spend $10–20.",
"languageCode": "en-US"
},
"references": {
"reviews": [
{
"name": "places/ChIJQ8nPjFW3j4ARcAaUg5pF-Ts/reviews/ChZDSUhNMG9nS0VJQ0FnSUN4aUtDbVJREAE",
"relativePublishTimeDescription": "12 months ago",
"rating": 5,
"text": {
"text": "Insanely good Pakistani food with the perfect ambiance! This food never disappoints AND THEY ARE OPEN TILL MIDNIGHT!!",
"languageCode": "en"
},
"originalText": {
"text": "Insanely good Pakistani food with the perfect ambiance! This food never disappoints AND THEY ARE OPEN TILL MIDNIGHT!!",
"languageCode": "en"
},
"authorAttribution": {
"displayName": "Anam Qureshi",
"uri": "https://www.google.com/maps/contrib/116859422317024657778/reviews",
"photoUri": "https://lh3.googleusercontent.com/a-/ALV-UjWbImVTop_1dl9FTSmec0x00000000-cc-rp-mo"
},
"publishTime": "2023-05-09T07:24:36Z"
},
. . .
]
}
}
},
{
"id": "ChIJETHw0F-7j4AR85V_I6OobCI",
"displayName": {
"text": "Hyderabad Dum Biryani",
"languageCode": "en"
},
"generativeSummary": {
"overview": {
"text": "Indian restaurant offering comfort food including vegetarian and vegan options.",
"languageCode": "en-US"
},
. . .
}
Request a generative area summary
Area summaries are generative summaries created by the LLM for the nearby area of a place. Area summaries can be important to users making a decision about a place.
For example, you are looking to book a hotel in a new city and while you can find a hotel that meets your requirements, you would also like to find out more about the area around the hotel. Rather than look at a map of the surrounding area, and then having to investigate each place on the map, you can instead generate an area summary for the hotel.
Area summaries are supported by Place Details (New), Text Search (New), and Nearby Search (New).
To include the area summary in the response, include in the field mask of the request:
areaSummary(Place Details)places.areaSummary(Text Search and Nearby Search).
The
areaSummary
field for a place in the response contains an array of up to four
contentBlocks.
The topic for each block is based on the type of the place:
For an EV charging station, defined as a place of type
electric_vehicle_charging_station, the topic blocks include:"overview","coffee","restaurant", and"store".For a place with a type defined by the Lodging category, and the specific place types
geocodeandpremise, the topic blocks include:"overview","activity","restaurant", and"store".
Example request for area summary of an EV charging station
This example Place Details request returns the area summary for a specific EV
charging station in Mountain View, identified by the place ID
ChIJC5b6QKiwj4ARgn2EwlnJnWU:
curl -X GET -H 'Content-Type: application/json' \ -H "X-Goog-Api-Key: API_KEY" \ -H "X-Goog-FieldMask: id,displayName,areaSummary" \ https://places.googleapis.com/v1/places/ChIJC5b6QKiwj4ARgn2EwlnJnWU
The response contains the areaSummary field, including the contentBlocks
array:
The
topicfield specifies the type of content block. Because the place in this example is an EV charging station, the four content blocks are"overview","coffee","restaurant", and"store".The
contentfield includes the summary for that topic, including the language code.The
referencesarray specifies the place resource names, in the formplaces/PLACE_ID, used to provide content for the content block.
{
"id": "ChIJC5b6QKiwj4ARgn2EwlnJnWU",
"displayName": {
"text": "EVgo Charging Station",
"languageCode": "en"
},
"areaSummary": {
"contentBlocks": [
{
"topic": "overview",
"content": {
"text": "Within a 4-minute walk, there are several dining and shopping options from this charging station, including Starbucks and Los Altos Taqueria.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJ-f-lEaiwj4ARB1IwfdLFfRY",
"places/ChIJhzHBsAe6j4ARvq9oi8u-bqQ"
]
}
},
{
"topic": "coffee",
"content": {
"text": "Starbucks offers Wi-Fi, light bites, and its signature roasts in a cozy space.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJ-f-lEaiwj4ARB1IwfdLFfRY"
]
}
},
{
"topic": "restaurant",
"content": {
"text": "Los Altos Taqueria serves classic Mexican dishes in a kid-friendly venue.\nHyderabad Dum Biryani specializes in vegetarian and vegan Indian fare, including lunch and dinner options.\nSushi Jin is a popular spot for sushi and other Japanese dishes, featuring a relaxed atmosphere.\nSuper tacos el conrro is a kid-friendly spot for Mexican fare.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJhzHBsAe6j4ARvq9oi8u-bqQ",
"places/ChIJETHw0F-7j4AR85V_I6OobCI",
"places/ChIJ-9pg_Qmxj4AR4deSLfZINW0",
"places/ChIJvcqEhTyxj4AR4ZdQevNxGZk"
]
}
},
{
"topic": "store",
"content": {
"text": "Safeway provides a range of groceries, along with a deli and a bakery.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJF7_iDaiwj4ARzdC4NlGzJFI"
]
}
}
]
}
}
Example request for area summary for a hotel
This example Text Search request requests hotels in Mountain View, CA:
curl -X POST -d '{
"textQuery": "Hotels",
"location_bias":{
"rectangle":{
"low": {
"latitude": 37.415,
"longitude": -122.091},
"high":{
"latitude": 37.429,
"longitude": -122.065}
}
},
"maxResultCount": 5
}' \
-H 'Content-Type: application/json' -H "X-Goog-Api-Key: API_KEY" \
-H "X-Goog-FieldMask: places.id,places.displayName,places.areaSummary" \
https://places.googleapis.com/v1/places:searchText
The response contains the areaSummary field, including the contentBlocks
array. The response for places of type hotel contains content blocks for the
topics "overview", "activity", "restaurant", and "store":
{
"places": [
{
"id": "ChIJ81X7bC63j4ARNJTuqMFBq-4",
"displayName": {
"text": "Shashi Hotel Mountain View, an Urban Resort",
"languageCode": "en"
},
"areaSummary": {
"contentBlocks": [
{
"topic": "overview",
"content": {
"text": "This area offers several attractions, including Cinemark Century Mountain View 16 and Computer History Museum. Charleston Park provides picnic spots and the Pear Theatre hosts performances.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJOYvCo1W3j4AR1LAifgk13rs",
"places/ChIJm7NJkla3j4AR8vR-HWRxgOo",
"places/ChIJo4za6vi5j4ARMua093RgGGA",
"places/ChIJ4V-yslC3j4ARE6I2cpvyOeU"
]
}
},
{
"topic": "activity",
"content": {
"text": "Cinemark Century Mountain View 16 offers XD and RealD 3D screens, plus comfortable seating and a snack bar.\nComputer History Museum showcases the evolution of technology with interactive exhibits.\nRyptic Room Escape Mountain View presents themed rooms where teams solve puzzles to escape.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJOYvCo1W3j4AR1LAifgk13rs",
"places/ChIJm7NJkla3j4AR8vR-HWRxgOo",
"places/ChIJ1_ZkCKqwj4ARPaLA4yiEtdI"
]
}
},
{
"topic": "restaurant",
"content": {
"text": "Cucina Venti Restaurant prepares Italian cuisine in a rustic setting with outdoor seating.\nRoger Bar and Restaurant provides cocktails on tap, along with an indoor-outdoor bar and dining area.\nBroma Restaurant serves Spanish-inspired tapas and shareable plates in a trendy atmosphere.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJC_3cdla3j4ARMgf4Pw8Irek",
"places/ChIJh1Y1MA23j4AREJz6Kxl75Uo",
"places/ChIJLRvDZtu3j4ARgoNrKZnNw8g"
]
}
},
{
"topic": "store",
"content": {
"text": "NASA Gear Store and Food Mart offer a variety of gifts and convenience store items.",
"languageCode": "en-US"
},
"references": {
"places": [
"places/ChIJh64-r0O3j4ARy1P41IH2j3k",
"places/ChIJ5ZJtB8u3j4ARR0gohzD8-ns"
]
}
}
]
}
},
...
]
}
Request contextual search results
The response to a Text Search (New) request contains an array of places that match the criteria of the input search string. Now you can request contextual search data in the response based on the input query to the API.
To support contextual summaries, the response from
Text Search (New)
contains a new array named contextualContents at the same level as the
places array in the response:
{
"places": [],
"contextualContents":[]
}
Each element in the contextualContents array, of type
ContextualContent,
is at the corresponding array location as the place in the places array. That
is, the element at contextualContents[0] corresponds to the place at
places[0].
Each array element contains the following array fields:
contextualContents.reviews: An array of up to three reviews.contextualContents.photos: An array of up to five photos.contextualContents.justifications: An array of justifications for the place, including:reviewJustification: Highlight a section of a user review that could be of interest to an end user.businessAvailabilityAttributesJustification: Includes attributes for a business that could be of interest to an end user.
To include the contextual search results in the response, include
contextualContents in the
field mask
of the Text Search (New) request:
curl -X POST -d '{
"textQuery": "Spicy vegetarian restaurant in Mountain View, CA",
"openNow": true,
"maxResultCount": 3
}' \
-H 'Content-Type: application/json' -H "X-Goog-Api-Key: API_KEY" \
-H "X-Goog-FieldMask: places.id,places.displayName,contextualContents" \
https://places.googleapis.com/v1/places:searchText
The response contains the contextualContents array for each place in the
response:
{
"places": [
{
"id": "ChIJETHw0F-7j4AR85V_I6OobCI",
"displayName": {
"text": "Hyderabad Dum Biryani",
"languageCode": "en"
}
},
. . .
],
"contextualContents": [
{
"reviews": [
{
"name": "places/ChIJETHw0F-7j4AR85V_I6OobCI/reviews/ChdDSUhNMG9nS0VJQ0FnSURSeEtMdHZ3RRAB",
"relativePublishTimeDescription": "a year ago",
"rating": 5,
"text": {
"text": "Very tasty! Rich spices and textures. We really appreciated
all of the vegan options. We shared samosas (great sauce!) , tofu 65
(spicy!!), gobi manchurian (less flavorful but a nice sweet offset to
the tofu 65) and veg pakoras (very crunchy and awesome flavor). We
enjoyed all of it, with the pakoras being the table's favorite.",
"languageCode": "en"
},
"originalText": {
"text": "Very tasty! Rich spices and textures. We really appreciated
all of the vegan options. We shared samosas (great sauce!) , tofu 65
(spicy!!), gobi manchurian (less flavorful but a nice sweet offset
to the tofu 65) and veg pakoras (very crunchy and awesome flavor).
We enjoyed all of it, with the pakoras being the table's favorite.",
"languageCode": "en"
},
"authorAttribution": {
"displayName": "Guillaume Riesen",
"uri": "https://www.google.com/maps/contrib/1131462893/reviews",
"photoUri": "https://lh3.googleusercontent.com/a-/ALVVXe4iwG_WXt1SjB1rJg=s128-c-c0x00000000-cc-rp-mo-br100"
},
"publishTime": "2023-04-18T02:48:02Z"
},
{
"name": "places/ChIJETHw0F-7j4AR85V_I6OobCI/reviews/ChdDSUhNMG9nS0VJQ0FnSUNKZzlhTG53RRAB",
"relativePublishTimeDescription": "10 months ago",
"rating": 5,
"text": {
"text": "This is a nice little spot to eat at. I'm a big fan of
curries, and they are spot on when it comes to flavor and spice. I highly recommend this place.",
"languageCode": "en"
},
"originalText": {
"text": "This is a nice little spot to eat at. I'm a big fan of curries,
and they are spot on when it comes to flavor and spice. I highly recommend this place.",
"languageCode": "en"
},
"authorAttribution": {
"displayName": "Paul Sacks",
"uri": "https://www.google.com/maps/contrib/1106671734435/reviews",
"photoUri": "https://lh3.googleusercontent.com/a/ACg8ocIAGguhz-EccsS9sVQ=s128-c-c0x00000000-cc-rp-mo-ba4-br100"
},
"publishTime": "2023-07-03T16:51:49Z"
},
. . .
],
"photos": [
{
"name": "places/ChIJETHw0F-7j4AR85V_I6OobCI/photos/Vt2DAKajwJfgC_I-baPKjSeSw-JAXY-tCZOBAe7pBxxIS0Vo1w23sCO9o5RzoxJ",
"widthPx": 1920,
"heightPx": 1080,
"authorAttributions": [
{
"displayName": "A Google User",
"uri": "//maps.google.com/maps/contrib/113209927655010465422",
"photoUri": "//lh3.googleusercontent.com/a-/ALV-UjV_8hbrtgycOLDgOqPRYhzJ0EJlBFyKo9OrmacOo3C5YvrtR4Q=s100-p-k-no-mo"
}
]
},
{
"name": "places/ChIJETHw0F-7j4AR85V_I6OobCI/photos/AUGGfZkJ4MMlLikE2EyP-b13GeFZ9YNzyjkjDMD15gyobIQtIYg1GFPy3BdQHAaA3iNwhk24vK212hZo",
"widthPx": 960,
"heightPx": 720,
"authorAttributions": [
{
"displayName": "Toast"
}
]
},
. . .
],
"justifications": [
{
"reviewJustification": {
"highlightedText": {
"text": "\"Not too spicy or oily.\"",
"highlightedTextRanges": [
{
"startIndex": 9,
"endIndex": 14
}
]
},
"review": {
"name": "places/ChIJETHw0F-7j4AR85V_I6OobCI/reviews/ChZDSUhNMG9nS0VJQ0FnSUR4djR2Z1B3EAE",
"relativePublishTimeDescription": "10 months ago",
"rating": 5,
"text": {
"text": "Not too spicy or oily. We have the mutton dum biryanis
and tandoori fish. Very flavorful. The food was served promptly.
Clean environment. A little small more like a hole in the wall.
We'd come back for sure.",
"languageCode": "en"
},
"originalText": {
"text": "Not too spicy or oily. We have the mutton dum biryanis
and tandoori fish. Very flavorful. The food was served promptly.
Clean environment. A little small more like a hole in the wall.
We'd come back for sure.",
"languageCode": "en"
},
"authorAttribution": {
"displayName": "Harpal Kochar",
"uri": "https://www.google.com/maps/contrib/1166232959/reviews",
"photoUri": "https://lh3.googleusercoEnzB1zdrNbR00000-cc-rp-mo"
},
"publishTime": "2023-06-18T06:12:41Z"
}
}
},
{
"businessAvailabilityAttributesJustification": {
"dineIn": true
}
}
]
},
. . .
]
}
Implementation guidance for automotive use cases
For driver safety, Google recommends rendering generative AI summaries in parked mode for automotive experiences. If rendering summaries in driving mode, on-screen information should be simple, clear, and concise to ensure that the driver's attention returns quickly to the road.
- No more than 3 lines
- 120 characters or 24 words or less
- Minimum front size of 24dp
Especially in automotive contexts, it's the developer's responsibility to render generative AI summaries in ways that are safe and compliant with all local laws and regulations.
Attributions
When you display reviews, photos, and generative AI summaries in your app, including both place and area summaries, you must ensure that you also display the required attributions.
Display attributions for generative summaries
When you display generative AI summaries in your app, including place and area summaries, you must also display the necessary attributions. This section describes how to display attributions based on how you display the generative AI summaries in your app.
In all cases, the attribution must use sentence casing. The attribution and Google logo must conform to all color, style and typeface standards. For more information on these standards, see Policies for Places API (New).
Generative AI summary displayed in the context of other Google data
In this case, the generative AI attribution can be displayed alone above or below the summary, relying on Google attribution of the content elsewhere.
For the attribution, use the text "Summarized by AI" as shown in this image:
Generative AI summary displayed in isolation
When a generative AI summary appears in isolation of any other content from Google Maps, it must include both generative AI attribution and attribution to Google separated by an interpunct. For the attribution, use the text "Summarized by AI" and the Google logo as shown in this image:
Generative AI summary repeated in a list
When repeating a generative AI place summary overview in a list of places, where all:
- List content is attributed to Google
- Summaries shown are AI generated
- Content is available from Google
The attribution should be displayed above or below the list on a single line. The AI attribution is left-aligned and the Google attribution is right-aligned:
If longer form generative AI description summaries are used in a list, each summary must have AI attribution ("Summarized by AI" with the info icon/button).
Display information about a generative AI summary in a dialog
Along with a generative AI summary, you must make a dialog available to end users that displays information about the summary, and a link to report issues. When using a generative AI description summary, you must include information about sources of the summary, called the references.
Enable access to this dialog by including an info icon (as an interactivity affordance) in conjunction with AI attribution (as depicted in the above examples) and making the AI attribution area and icon respond to a press/click.
For generative place summaries, display:
| About | AI place summaries are generated from Google reviews and other data. |
| Report an Issue | To report an issue to Google about the content of a summary, follow this link. |
| References | Source reviews |
For example:
For generative area summaries, display:
| About | AI area summaries are generated from Google reviews and local data. |
| Report an Issue | To report an issue to Google about the content of a summary, follow this link. |
