Who uses wheresthe.ai, and how

Eight real jobs, each with an example scenario, the exact steps on this site, and where it fits in the work. The people are hypothetical; every rule, date and number is real and links to its record.

wheresthe.ai does thisgives you a head startstill your team or counsel

Analysts, strategists and the curious

Getting up to speed on military AI in an afternoon

Strategy and policy analysts, consultants, investors, students and anyone who needs to understand an AI story fast, beyond the headline.

Example scenario. Arjun is a strategy consultant in Delhi. A client in defence electronics asks on Monday morning what Project Meridian means for them. He has read one headline and has until the afternoon call.

Without this site: Twenty browser tabs of news coverage, no way to tell which claims came from an official source, and no sense of what led here or what other countries are doing.

  1. Search for "Meridian" (press / on any page). The top result is the curated line Project Meridian and the race to govern military AI: ten stops from the EU AI Act's military exclusion to the announcement itself, each with a one-line note on why it matters.
  2. Open the Project Meridian record. It is marked Reported because the Department of War memo could not be opened, with the coverage it relies on listed under Source.
  3. Follow its Line of Thought: the January 'AI-first' strategy led here; under "Same story elsewhere" sit the UK's Strategic Defence Review, the EU's defence White Paper and the UN vote on autonomous weapons. Each click adds a stop to his own line at the top of the page.
  4. Copy the link to his line of thought and paste it into the client note, so the client can retrace every step and source. For the supplier angle he adds Who supplies the AI-first military?.

What they walk away with: A sourced, dated briefing on what happened, why, and how other governments are responding, with a shareable trail instead of a pile of tabs.

Compliance and AI governance

Launching an AI hiring tool in five new markets

AI governance leads, compliance officers and in-house lawyers at companies that build or buy AI.

Example scenario. Priya leads AI governance at a Bengaluru HR-tech company. Its CV-screening product launches next year in the EU, New York City, Illinois, Colorado and California. Her CEO wants one page by Friday: what must we do, where, and by when?

Without this site: Five legislature websites, the EU Official Journal, a stack of law-firm alerts, and no single place that shows which dates moved this year.

  1. Open the Rules page filtered to employment: 10 instruments. Five of them apply to her launch markets.
  2. Read each record for what it requires and when: NYC Local Law 144 (independent bias audit before use, in force since 5 Jul 2023), Illinois HB 3773 (in force 1 Jan 2026), Colorado SB 26-189 (applies 1 Jan 2027), California SB 947 (from 1 Jul 2027) and the EU AI Act, where hiring is high-risk and, after the Digital Omnibus, those duties start 2 Dec 2027.
  3. Line the markets up side by side in Compare to see which obligations repeat (bias testing, notices, human review) and build one control set instead of five.
  4. Add every date to the team calendar from the deadline feed, and turn on alerts for each jurisdiction so a moved date doesn't surprise her.

What they walk away with: A dated obligations checklist for five markets, each line linked to the official text, ready for outside counsel to review instead of research from scratch.

Founders and product teams

Shipping an AI companion app in the US and China

Founders, product managers and trust-and-safety leads deciding what a feature must do before launch.

Example scenario. Leila is a founder in San Francisco. Her AI companion app launches in California and New York this quarter, with a China version planned through a local partner. She wants the legal must-haves in the product spec now, not after launch.

Without this site: Asking a lawyer an open question at an hourly rate, or finding out from an app-store rejection or a regulator.

  1. Search the Rules page for "companion". Four rules come back: California SB 243 (AI disclosure and suicide-risk protocols, in force 1 Jan 2026), New York's AI companion law (in force since 5 Nov 2025), California SB 1119 (child-safety risk assessments from 1 Jul 2027) and China's companion AI measures (in force 15 Jul 2026).
  2. For China, read the Generative AI Measures: public services need a CAC filing. Search the China registry to check her partner's model is one of the 1,115 filed services.
  3. Turn the obligation tags into spec items: an "you are talking to AI" disclosure, a crisis-referral flow, minor protections and content labels.
  4. Put 1 Jul 2027 (SB 1119) on the roadmap, and set alerts for California, New York and China.

What they walk away with: A product requirements list per market with launch-blocking items flagged, written before engineering starts.

Policy research

Writing a comparative paper on deepfake labelling

Think-tank fellows, academics, students and civil servants who compare how countries govern AI.

Example scenario. Ananya is a research fellow at a Delhi think tank. She is writing a paper comparing India's new deepfake-labelling rules with China, the EU and South Korea, and needs every claim cited to a primary source.

Without this site: Weeks of collecting gazette notifications, translating Chinese measures, and checking which drafts became law.

  1. Filter the Rules page to labelling AI content: 15 binding instruments across 8 jurisdictions, plus drafts and guidance.
  2. Start with India: the IT Rules amendment on synthetic media (in force 20 Feb 2026) and the draft Second Amendment still in consultation.
  3. Put India, China and the EU side by side for the comparison table, then add South Korea.
  4. Copy the ready-made citation from each rule page, download all rules as CSV for the annex, and embed a chart with credit.

What they walk away with: A comparison table and annex with a primary-source citation for every line, so the paper can focus on the argument.

Journalists

Reporting on AI rules without getting the dates wrong

Technology and policy reporters, editors and newsletter writers.

Example scenario. Tom is a technology reporter at a business daily in London. On Monday morning he needs what changed in AI rules last week, and a fast check on whether the law a source mentioned is actually in force.

Without this site: Trusting a press release or an old explainer. The EU AI Act timeline changed in July 2026, and plenty of coverage still uses the old dates.

  1. Read "What changed" on the home page or subscribe to the RSS feed. Last week's items included California's SB 947, signed 30 Sep 2026.
  2. Before quoting a status, open the record. The EU AI Act is partly in force, and its high-risk duties now start 2 Dec 2027, not Aug 2026.
  3. Use a chart: the China registry shows filings rising from 302 in Jan 2025 to 1,115 now. Every chart has an embed code with credit.
  4. Get the Monday brief by email so the scan takes five minutes.

What they walk away with: Correct statuses and dates with a source link, and a chart for the story.

Investors and analysts

Checking a Chinese AI company's regulatory claims

VC and PE analysts, corporate development and partnership teams, China-watchers.

Example scenario. Mei is an analyst at a Singapore venture fund looking at a Chinese AI startup whose deck says its model is "approved in China". She needs to know what that means and whether it is true.

Without this site: Chinese-language PDFs from cac.gov.cn, published in batches, with no search and no English names.

  1. Search the China registry by company or model, in English or Chinese. Searching "Moonshot" finds Kimi, filing number Beijing-MoonShot-20231016, listed in the 2 Apr 2024 bulletin.
  2. Check the kind. "Filed" means the model passed CAC security assessment. "Registered" means the app calls someone else's filed model, which is a weaker claim.
  3. Read the context: filings rose from 302 in Jan 2025 to 1,115. A filing is a licence to operate, not a moat.
  4. Skim China's rulebook for duties the company must keep meeting, such as the AI content labelling measures.

What they walk away with: A verified line in the investment memo: filed or registered, with the official filing number.

Health tech and regulatory affairs

Benchmarking medical AI clearances before an FDA submission

Regulatory affairs leads, health-tech founders and med-tech investors.

Example scenario. Dr Arjun Rao is the regulatory lead at a radiology AI startup in Pune, preparing its first FDA 510(k) and planning an EU launch. He wants to see who from India has cleared the FDA, and for what.

Without this site: The FDA list doesn't say where companies are based, so finding Indian clearances means checking applicants one by one.

  1. Open the FDA registry and filter to India: 13 devices, 12 of them in radiology.
  2. Filter by Radiology to see the field: 1,230 of 1,614 AI devices (76%). Each row links to its FDA record, a starting point for finding predicate devices.
  3. Check the rules: the EU AI Act treats AI in medical devices as high-risk, with duties now due 2 Aug 2028, and India's ICMR guidelines cover clinical AI ethics.
  4. Watch for new clearances from competitors with each FDA list update.

What they walk away with: A competitor and predicate shortlist with FDA links, and a dated view of EU and Indian obligations.

Banks and financial services

Rolling out an AI credit model at an Indian lender

Model risk, compliance and data protection teams at banks, NBFCs and fintechs.

Example scenario. Kabir leads model risk at a Mumbai NBFC rolling out an AI underwriting model. The board wants to know which rules apply, what is still in draft, and what lands in the next 12 months.

Without this site: Tracking RBI, SEBI and MeitY circulars separately, and missing that a draft has opened for comments.

  1. Filter to India and finance: 3 instruments, including the RBI FREE-AI framework (Aug 2025) and the draft RBI model risk guidance, out for consultation since 24 Jun 2026.
  2. Add the cross-cutting law: the DPDP Act and Rules. Consent-manager rules apply from 13 Nov 2026 and the main obligations from 13 May 2027.
  3. If the group has a SEBI-regulated arm, read SEBI Regulation 16C: the firm stays responsible for any AI tool it buys.
  4. Turn on alerts for India so the board hears the day the RBI draft becomes final.

What they walk away with: A board note listing what applies now, what is in draft and the dates ahead, each with its official circular.

Building something on this data?

Developers and AI agents can use the free MCP server, JSON API and Markdown pages. No key needed.

See the developer page