TL;DR: Google Knowledge Graph News for November 2026 and What It Means for Your Startup
Google Knowledge Graph news, November, 2026 carries one blunt message for you: AI agents now read Google's entity data directly. If your startup is missing from that data, it is also missing from their answers. Founder Violetta Bonenkamp (Mean CEO) explains why checking your entity status takes one afternoon and helps you show up in AI-generated search answers.
- What changed: On November 2, 2026, APIs.io labelled the Knowledge Graph Search API as freemium and self-serve. It also gave the API an "agentic access" profile for AI agents. The API itself stays the same. It offers one read-only call, entities:search, which returns about 8 schema.org fields in JSON-LD.
- Why you should care: Investors, competitors and their AI research tools can audit your entity footprint in seconds, for free. A missing or duplicated entity makes your company look like it barely exists during due diligence.
- How to check, with no code:
- Create a Google Cloud project and turn on the API
- Query your company, founder and product names
- Save your kg:/m/ or kg:/g/ entity ID
- Rerun the same queries every month
- How to fix gaps: You cannot edit the graph through the API. Publish Organization and Person JSON-LD markup and use one spelling of your name everywhere. Collect independent citations, then claim your panel using this knowledge panel guide. Test product names before launch, because names like "Boris" or "Atlas" get lost among thousands of existing entities.
Pair your audit with an entity-led content strategy so machines link your founders, products and brand correctly. An entity check costs you one afternoon, while staying invisible to the tools that answer your customers' questions costs far more. Your first query this week is the cheapest move you can make.
Check out other fresh startup news and trends that you might like:
WordPress Security News | November, 2026 (STARTUP EDITION)
Google Knowledge Graph news for November 2026 looks quiet on the surface, and that quiet hides the most useful story of the month for anyone building a company. The Knowledge Graph Search API, the only official public door into Google’s private database of people, places, organizations and things, still does exactly what its documentation promises. It lets you find entities, it uses standard schema.org types, and it returns results in JSON-LD. What changed is the context around it. On November 2, 2026, the API catalog APIs.io regenerated its profile of the service, labelling it FREEMIUM and SELF-SERVE, with no trial and no gated access, and listing an “agentic access” property that maps which operations an AI agent can call without a human in the loop.
I am Violetta Bonenkamp, better known as Mean CEO. I run CADChain, a deeptech company working on IP protection for CAD and 3D data, and Fe/male Switch, a startup game and online incubator for women founders, in parallel. I have spent over 20 years working across Europe, and my academic background in linguistics shapes how I read this news. Language is an interface between humans and machines. The Knowledge Graph is where Google decides what your company IS, who your founders ARE, and whether your product deserves a box on the right side of a search result. If you are a founder, freelancer or business owner, this month’s news tells you something blunt: machines are starting to read Google’s entity data directly, and if you are missing from it, you are missing from their answers.
Let’s break it down. Below you will find what happened, what the API returns, a step-by-step guide to checking your own entity status, the mistakes I see founders make over and over, and my honest take on where this is heading.
What is the Google Knowledge Graph, in plain words?
The Google Knowledge Graph is a knowledge base that Google added to Search on May 16, 2012. It stores facts about real-world things, called entities, and the relationships between them. An entity can be a person (a founder), an organization (your startup), a place (your office city), a creative work (a book or a film) or an event. Each entity is a node in the graph, and the connections between nodes are the facts: “Violetta Bonenkamp” is the CEO of “CADChain”, which is located in “the Netherlands”.
Google uses this graph to interpret queries, tell apart people and companies with similar names, and fill knowledge panels, the information boxes that show up next to search results. When Google announced an update to the graph in 2020, it said the Knowledge Graph held 500 billion facts about 5 billion entities. That number is now several years old, so treat it as a floor, not a ceiling. The graph is private. You cannot download it or edit it directly, and that is exactly why the public API matters.
Which “knowledge graph” are we talking about?
The term gets confusing fast, so here is a quick disambiguation before we go further:
- Google Knowledge Graph: Google’s private, public-facing entity database that powers Search features.
- Knowledge Graph Search API: the read-only developer interface that lets you query entities in that database.
- Enterprise Knowledge Graph: a Google Cloud product that helps companies build a private graph of their own records (customers, suppliers, people) and reconcile duplicates.
- Knowledge panel: the visible box in search results that displays data pulled from the graph.
- A knowledge graph (generic): any graph database of entities and relations, such as Wikidata or DBpedia.
What happened in Google Knowledge Graph news in November 2026?
No splashy product launch, no new endpoint, no rename. The developments are structural, and structural news is the kind founders usually ignore until it bites them. Here is what I tracked this month.
- APIs.io refreshed its access model profile on November 2, 2026. The Google Knowledge Graph Search profile on APIs.io now classifies the API as a hosted service you call directly, with a freemium pricing model, self-serve signup, no trial and no gating. The catalog entry was created on March 13, 2026, modified on May 19, 2026, and its delivery model data was generated on August 28, 2026.
- The API is documented for AI agents. The API Evangelist third-party profile of the Knowledge Graph Search API on GitHub lists an “Agentic access” property describing which operations an agent can call, which ones act, and which need a human in the loop. It also publishes a JSON Schema for the entity model, a JSON-LD context, three request collections (Postman and Open Collection), two Spectral governance rulesets, plus pages on authentication, rate limits and pricing.
- Google’s official documentation stays stable. The official Google Knowledge Graph Search API documentation still describes three typical use cases: getting a ranked list of prominent entities matching certain criteria, predictively completing entities in a search box, and annotating or organizing content with Knowledge Graph entities.
- The Enterprise Knowledge Graph route keeps the same output format. The Google Cloud Enterprise Knowledge Graph Search API documentation shows requests authenticated with a bearer token from the gcloud CLI and confirms the response is a list of entities in JSON-LD, “compatible with schema.org schemas with limited external extensions.”
- Old gaps are still open. A Stack Overflow thread on which schema.org types the Knowledge Graph API supports, first asked in June 2019 and now over seven years old, still documents that the API does not return results for the
ComputerLanguagetype and has no Python programming language entity, while DBpedia and Wikidata both do.
My read: the headline of November 2026 is that a 2015-era API has quietly become machine-catalogued infrastructure for AI agents. A tool once used mostly by SEO specialists and curious developers is now packaged with schemas and governance rules so that autonomous software can query it. That changes who is asking Google “who is this company?”, and how often.
What does the Knowledge Graph Search API actually return?
The API exposes a single operation, entities:search. You send a text query or an entity ID, and you get back a JSON-LD document. Inside it sits an itemListElement array, and each member carries a result object plus a resultScore. According to the Knowledge Graph Navigator guide to how the Google Knowledge Graph works, the result object exposes a small, fixed set of fields.
- @id: the canonical identifier for the entity, prefixed with
kg:. - name: the entity’s name.
- @type: a list of schema.org types that match the entity, such as Person, Organization or Thing.
- description: a short label, like “Martial Artist” or “Software company”.
- detailedDescription: a longer text, often with a source URL and license.
- image: an image related to the entity.
- url: the official website of the entity.
- resultScore: an indicator of how well the entity matches your query.
A classic response, shared by developer Uwe Fetzer in a SAP Community post on the Google Knowledge Graph Search API, schema.org and JSON-LD, returns the entity kg:/m/015lhm with the name “Chuck Norris”, the types Person and Thing, and the description “Martial Artist”. That same 2016 post shows the API wrapped in ABAP code, which proves a point I love making to founders: you can call this API from almost any stack, including no-code automation tools that send simple HTTP requests.
What are MIDs and why should a founder care?
Every entity carries an opaque identifier. Older ones look like /m/0dl567, a Freebase machine ID (MID) inherited from Freebase, the community database Google acquired and later shut down. Newer entities start with /g/, such as /g/11b6t0kqhx. Neither string encodes the entity’s name or type, and both stay stable when the name changes. That stability is the whole point.
For a founder, your MID is the closest thing you have to a machine passport. If your startup rebrands, the ID can survive the rename. If you find two IDs for your company, Google thinks you are two different things, and that splits your authority across both.
Why should entrepreneurs and freelancers care about Knowledge Graph news at all?
Here is why. Search is shifting from “ten blue links” to answers, and answers are built from entities. As the Search Engine Journal guide to the Knowledge Graph Search API for SEO explains, entities let Google move past simple string matching toward its goal of becoming an answer engine. Google surfaces graph data in knowledge panels, knowledge cards and featured snippets, and structured data on websites can influence what gets pulled into the graph.
Numbers that should make you slightly uncomfortable
- 500 billion facts about 5 billion entities: the size Google reported for its Knowledge Graph in 2020. Your company is competing for recognition inside that crowd.
- 14 years: the Knowledge Graph has existed since May 2012. Competitors who started building entity signals early have a long head start.
- 1 operation: the public API offers exactly one call, entities:search. Google gives you a peephole, never a door.
- 8 fields: the API returns a fixed handful of fields per result. Relationships, founders, funding rounds and product lines stay hidden.
- Over 7 years: how long the Stack Overflow question about missing programming language entities has remained open. Gaps in the graph can persist for years.
- 0 trial, 0 gating: per the November 2026 APIs.io refresh, anyone can sign up self-serve. Your competitors can check your entity status today, for free.
One more detail from Search Engine Journal is worth your attention: the API is independent of the user’s location. A regular Google search is personalized by country, language and history. The API gives you a cleaner view of how Google’s database sees an entity, which makes it a far better audit tool than typing your brand name into your own browser.
How do you check if your startup exists in the Google Knowledge Graph?
This is the how-to part. You need no engineering team. I follow a simple principle: default to no-code until you hit a hard wall. You can run every step below from a browser or a no-code automation tool.
- Create a Google Cloud project. Sign in to the Google Cloud Console and create a new project with an obvious name, like “entity-audit”.
- Switch on the Knowledge Graph Search API. Find it in the API library and turn it on for your project.
- Generate an API credential. Create one in the Credentials section and restrict it to the Knowledge Graph Search API so nobody can misuse it for other services.
- Send your first query. Call
https://kgsearch.googleapis.com/v1/entities:searchwith the parametersquery=YourCompanyName,types=Organizationandlimit=10, then append your API credential as the final parameter, as shown in Google’s sample requests. - Read the results carefully. Look for your company name, check the
@type, the description and theurl. Is the website correct? Is the description current? - Record your @id. If you appear, copy the
kg:/m/...orkg:/g/...value into a spreadsheet. This is your entity ID. - Compare the resultScore. If a competitor or a namesake ranks above you for your own brand name, you have an ambiguity problem.
- Repeat for people and products. Query your founder names with
types=Personand your product names with no type filter. - Schedule a monthly check. Set up a recurring automation that runs the same queries and logs changes. Entity data shifts quietly, and nobody sends you a notification.
What should you do if you are not in the graph?
Do not panic, and do not buy “knowledge panel packages” from strangers. Work on the signals Google reads. As WordLift notes in its guide to the Google Knowledge Graph and updating the Knowledge Panel, information in the graph is extracted from multiple sources, including structured data encoded in web pages using Microdata and JSON-LD.
- Publish Organization markup in JSON-LD on your homepage and About page, with your legal name, logo, founding date, founders and
sameAslinks to your official profiles. - Use one name everywhere. “Fe/male Switch”, “Female Switch” and “FemaleSwitch” look like three entities to a machine. Pick one canonical spelling.
- Build Person entities for founders. Add Person markup on founder bio pages and link them to the Organization with
founderproperties. - Get cited by independent sources. Press coverage, award listings, accelerator pages and conference speaker profiles give the graph corroborating facts.
- Claim your knowledge panel once it appears. Google lets verified representatives suggest edits to panels, which is the closest you will get to editing your entry.
A real-world example from my own work
When I check entity data for my own ventures, I query the company name, my name, and the product names separately. CADChain has products like Boris for Inventor, an Autodesk Inventor plugin, and BlendedBoris for Blender. A product named “Boris” is a perfect example of an ambiguity trap. A machine sees thousands of Borises: politicians, writers, characters. Without strong structured data tying the product to CADChain and to CAD workflows, an AI agent asking the graph “what is Boris?” will never land on our plugin. Naming decisions are entity decisions, and most founders make them without ever running a single API query.
How can founders use the Knowledge Graph Search API beyond vanity checks?
Checking your own name is step one. The real business use starts when you treat the API as a free, structured reference layer for your product and operations. Here are uses I recommend to small teams and solo founders.
- Smart search boxes. Use the API’s predictive completion to suggest real entities as users type. A recruiting startup can autocomplete company names; a travel blog can autocomplete cities.
- Content tagging. Annotate blog posts or course lessons with entity IDs. At Fe/male Switch, a quest about pitching to investors could be tagged with entities for specific accelerators, so related content connects automatically.
- CRM enrichment. When a lead types “Siemens” into a form, the API returns a canonical entity with a description and official URL. Fewer typos, cleaner records.
- Competitor mapping. Query your category terms and see which companies Google already treats as prominent entities. That list is your real competitive set in the eyes of machines.
- Grounding AI agents. If you build AI assistants, let them confirm names against the graph before writing about a person or company. This reduces confident nonsense about entities that do not exist.
- Entity reconciliation for B2B data. The Google Cloud walkthrough of Enterprise Knowledge Graph and the Entity Reconciliation API shows how to merge duplicate records of organizations, local businesses and people, then use the returned
cloud_kg_midto query the Knowledge Graph Search API for more details.
Here is a quick scenario. A freelance consultant serving manufacturing SMEs keeps a messy client list with “Philips”, “Philips NV”, and “Royal Philips” as three rows. Running each through the API returns the same entity ID, so she merges them in five minutes. That same list now feeds a newsletter tool, a proposal generator and a market report, all pointing at the same clean entity.
What are the most common Knowledge Graph mistakes founders make?
I mentor founders through Fe/male Switch and talk to engineers through CADChain, and the same errors show up in both worlds. Avoid these.
- Treating the API as an editing tool. It is read-only. You cannot push data into the Google Knowledge Graph through it. Anyone promising otherwise is selling fiction.
- Expecting relationships in the response. The API returns a fixed set of fields. Funding, founders, subsidiaries and product lines are not exposed, even when Google knows them.
- Judging your status from a personalized search. Your browser shows you a version of Google shaped by your location and history. Use the location-independent API.
- Inconsistent naming across the web. Different spellings on LinkedIn, Crunchbase, your website and press releases create competing signals.
- Using niche schema types the API ignores. As the Stack Overflow thread shows, some schema.org types return nothing. Stick with widely supported types like Organization, Person, Product, Event, Book and Place.
- Exposing your API credential in front-end code. Put calls behind a server or a no-code backend, and restrict the credential to this one API.
- Ignoring rate limits and quotas. The API is freemium with published request limits. Cache results instead of querying the same entity on every page load.
- Naming products after common words. A product called “Atlas” or “Nova” will drown among thousands of existing entities. Run an API query before you print the logo.
- Forgetting your founders. Investors, journalists and AI tools check people as well as companies. A founder without a Person entity looks thinner than they are.
What does Mean CEO think this news says about AI search and small businesses?
Now for the provocative part. I have three takes, and some readers will disagree with them.
Take 1: Your website is a draft and your entity is the published version
Founders spend months polishing landing pages. Then an AI assistant summarizes their company using three facts from an entity record they never looked at. When AI agents call APIs like entities:search, they read the compressed version of you. “If the machine cannot name you, it cannot recommend you.” I say this to every founder in our incubator, and I mean it literally.
Take 2: Entity gaps are a systems problem, especially for women founders
I often say that women do not need more inspiration; they need infrastructure. The Knowledge Graph reflects who already gets press coverage, award listings, conference stages and Wikipedia-style citations. Groups that get fewer of those signals end up thinner in the graph, and thinner entities get fewer mentions in AI answers. That loop compounds quietly. The fix is practical: structured data on your site, consistent naming, documented awards and speaker pages. Motivation speeches will not create a single MID.
Take 3: Entity hygiene should be invisible, the way IP protection should be
At CADChain, our philosophy is that protection and compliance should be invisible. Engineers should not have to become IP lawyers to stay safe, and founders should not have to become semantic web specialists to be findable. Website builders, CMS platforms and startup tools should generate clean Organization and Person markup by default. Until they do, you need to treat entity hygiene as a recurring task on your operations checklist, right next to bookkeeping and backups.
There is one more uncomfortable point. The APIs.io catalog describes the API as freemium and self-serve, and that openness cuts both ways. Your competitors, investors and their AI research agents can audit your entity footprint in seconds. During due diligence, a missing or confused entity looks like a company that barely exists, whatever your traction slide says.
Frequently asked questions about Google Knowledge Graph news and the API
Is the Knowledge Graph Search API free?
According to the November 2, 2026 APIs.io profile, the access model is freemium with self-serve signup and no trial. You need a Google Cloud project and an API credential, and published rate limits apply.
Can I add my company to the Google Knowledge Graph through the API?
No. The API is read-only. You influence the graph indirectly through structured data, consistent naming, independent citations and, once you have a knowledge panel, verified suggestions to Google.
What format does the API return?
It returns JSON-LD compatible with schema.org types, with a few Google-specific extensions such as resultScore and detailedDescription.
What is the difference between /m/ and /g/ IDs?
IDs starting with /m/ come from the Freebase era, while /g/ IDs belong to newer entities. Both are stable, opaque identifiers, and the API shows them with a kg: prefix.
Is the Enterprise Knowledge Graph the same as the Google Knowledge Graph?
No. Enterprise Knowledge Graph is a Google Cloud product for building and reconciling your own private entity data. It can link your records to Google Knowledge Graph IDs, which you can then look up with the search API.
What should founders do next after reading November 2026 Knowledge Graph news?
The November 2026 story is simple once you strip away the jargon. Google’s entity API has not changed much, and the world around it has. It is now catalogued, schema-documented and labelled for agent access, which means software, as well as people, keeps asking Google who you are. Founders who check and clean their entity data now will show up in tomorrow’s AI answers. Founders who wait will discover the problem when an investor’s research tool describes their startup wrongly or skips it entirely.
Next steps, in order of effort:
- This week: run the nine-step audit above for your company, your founders and your top product.
- This month: fix naming inconsistencies across your website, LinkedIn, directories and press kit, and publish Organization and Person JSON-LD.
- This quarter: collect independent citations through awards, accelerator listings, guest articles and speaker pages.
- Every month after that: automate a recurring API check and log changes in a spreadsheet.
- Before your next launch: query every candidate product name in the API before you commit to it.
Treat this like a game, the way I treat every startup challenge: collect information and assets faster than your competitors, and make the cheap moves first. An entity audit costs you an afternoon. Being invisible to the machines that answer your customers’ questions costs you far more.
People Also Ask:
What is a knowledge graph in layman terms?
A knowledge graph is a giant web of facts. Each dot in the web is a real thing, like a person, place, company, or idea. The lines between the dots show how those things connect. "Paris is the capital of France" and "Tom Hanks starred in Forrest Gump" are two such links. Search engines use this map to figure out what you mean, not just match the words you type.
How do I claim my Google knowledge panel?
Search for your name or organization on Google and find the knowledge panel. At the bottom of the panel, click "Claim this knowledge panel." Sign in with a Google account connected to an official website or profile, such as your YouTube channel, social media account, or Search Console property. Then follow the verification steps, which may ask for a screenshot or ID. Once Google approves the claim, you can suggest edits to the panel. Local businesses manage their panels through Google Business Profile instead.
Does Google have a graph maker?
Google doesn't offer a standalone graph maker, but Google Sheets lets you build charts for free. Select your data, then click Insert > Chart, and pick a line, bar, pie, or scatter chart. You can add these charts to Google Docs and Google Slides. Developers can also use the Google Charts library to put interactive charts on web pages. These tools have nothing to do with the Google Knowledge Graph, which is a database of facts rather than a charting tool.
What are the top 5 graph databases?
Five widely used graph databases are:
- Neo4j: Often ranked first and known for its Cypher query language.
- Amazon Neptune: A managed graph service on AWS.
- ArangoDB: A multi-model database that handles graphs, documents, and key-value data.
- TigerGraph: Built for fast analytics on very large graphs.
- Azure Cosmos DB: Supports graph data through its Gremlin API.
JanusGraph and Memgraph are also popular choices. Rankings shift from year to year.
What is the Google Knowledge Graph API?
The Knowledge Graph Search API is a free Google tool that lets developers look up entities stored in Google's Knowledge Graph. It returns results in JSON-LD format using schema.org types. People use it to:
- Get a ranked list of entities that match a search term.
- Auto-complete entity names in a search box.
- Tag or organize content by entity.
You need API credentials from the Google Cloud Console to use it. The API is read-only and returns only part of what Google stores. It does not give access to the full graph.
How can I search the Google Knowledge Graph?
The simplest way is a regular Google search. Knowledge panels and instant answers come straight from the graph. For direct lookups, send a query to the Knowledge Graph Search API. It returns matching entities with names, types, descriptions, and machine IDs that start with /m/ or /g/. Several free third-party tools put a search box on top of this API, so you can check whether a person or brand has an entity entry without writing code.
What is the difference between the Knowledge Graph and a Knowledge Panel?
The Knowledge Graph is the behind-the-scenes database where Google stores entities and the relationships between them. A Knowledge Panel is the info box you see in search results. It shows on the right side of the page on desktop and near the top on mobile, and it pulls its facts from the graph. The graph is the engine and the panel is the display. An entity can exist in the graph without Google showing a panel for it.
When did Google launch the Knowledge Graph?
Google launched the Knowledge Graph in May 2012 with the slogan "things, not strings." At launch it held about 500 million objects and 3.5 billion facts about them. By 2020, Google reported more than 500 billion facts about 5 billion entities. The graph keeps growing as Google pulls in data from websites, licensed sources, and public databases like Wikidata.
How do I get my business into the Google Knowledge Graph?
Follow these steps:
- Create and verify a Google Business Profile.
- Add Organization or LocalBusiness schema markup to your website, with "sameAs" links to your social profiles.
- Use the same business name, address, and phone number on every directory and listing.
- Earn mentions on trusted sites such as news outlets and industry publications.
- Add an entry to Wikidata if your business meets its notability rules.
These steps help Google treat your business as a confirmed entity.
Why does the Google Knowledge Graph matter for SEO?
Google now organizes search around entities instead of plain keywords. If Google knows your brand as an entity, you have a better chance of earning a knowledge panel, rich results, and mentions in AI Overviews and Gemini answers. Google uses the graph to check facts in those AI answers. The graph also answers many questions directly on the results page. SparkToro research found that about 60% of searches end without a click. That makes it important for brands to have accurate details in the graph, since those details often reach users before they ever visit a website.
FAQ on Google Knowledge Graph News and Entity Visibility in 2026
How does the Google Knowledge Graph shape AI Overviews and Gemini answers?
Generative answers rely on verified entities so they do not name the wrong company or person. Clean entity data raises the odds that AI Overviews and Gemini describe you accurately. Pair schema markup with consistent naming, and test branded prompts monthly. Explore AI SEO for startups.
Do founders need a Wikipedia article to earn a Google Knowledge Panel?
No. Wikipedia helps, but many startups earn panels through a sourced Wikidata item, schema markup and trusted profiles on LinkedIn, Crunchbase and media sites. Never write your own Wikipedia page, because editors usually delete conflict-of-interest edits. Read the Google Knowledge Panel guide for founders.
How does Wikidata feed the Google Knowledge Graph?
Wikidata offers machine-readable statements with stable Q-IDs that Google can match against other sources. A startup item needs independent references and correct properties like founder, inception and official website. Then add your Wikidata URL to the sameAs links in your Organization markup. See how Wikipedia and Wikidata build startup credibility.
Which pages should carry mainEntityOfPage markup?
Use it on your About page for the Organization, on founder bios for each Person, and on product pages for each Product or SoftwareApplication. One URL should describe one dominant entity. Mixed signals across pages blur how machines classify your startup. Apply MainEntityOfPage schema correctly.
How do I build an entity-led content plan instead of a keyword plan?
List the entities you want to own: your brand, founders, products and signature methods. Give each one a canonical page, a consistent definition and schema markup, then link supporting articles to those hubs. Track knowledge panel appearances alongside keyword rankings. Plan an entity-led content strategy.
How should international startups handle multilingual entity descriptions?
Google stores names and descriptions per language, so audit your entity with the API's languages parameter for each market. Keep the brand name identical everywhere, translate only the descriptor, and add hreflang tags and localized Organization markup. Mismatched translations can split recognition across countries. Audit your multilingual entity signals.
How can I open my entity directly in Google Search using its ID?
Copy the ID from the API's @id field and drop the kg: prefix. Then add it to google.com/search?kgmid=, so /g/11b6t0kqhx becomes google.com/search?kgmid=/g/11b6t0kqhx. This shows the exact panel tied to that entity, which helps you spot duplicate entities. Find your Google Knowledge Graph ID.
Which lesser-known API parameters make entity audits more accurate?
Beyond query, types and limit, the API accepts ids to fetch specific entities, languages to check localized descriptions, and prefix=true for partial, autocomplete-style matches. Add indent=true for readable JSON while testing. Combining types with languages reveals gaps that most founders miss. Check the Knowledge Graph Search API documentation.
How can I give an AI agent safe access to the Knowledge Graph Search API?
Wrap entities:search in a server-side tool so the agent never sees your credential. The endpoint is read-only, so the calls themselves carry little risk. Still, cache responses, cap requests per session, and have a human review anything the agent publishes about real people. Review the agentic access profile on GitHub.
How do I update an existing Google Knowledge Panel?
Search your brand, select "Claim this knowledge panel," and verify through an official account linked to your website or social profiles. Once verified, you can suggest edits to the featured image, description and profile links. Cite public sources Google can confirm. Understand how Knowledge Panel updates work.
