# Raghav R Handa

> Raghav R Handa is a lawyer-turned-builder in New Delhi, India, and the founder of SafeLaunch. He fixes apps built with AI tools such as Lovable, Bolt, Base44, Cursor, Replit and Claude Code, whatever state they are in, and makes them live, secure and legal. He does not practise law or offer legal services through this site; it is general information, not legal advice.

## Main pages

- [Home](https://raghavrhanda.com/): Raghav R Handa, founder of SafeLaunch: the story from a screwdriver at three to five years of law, a legal technology company and the checker he built, the things he has built, and his writing.
- [Services and prices](https://raghavrhanda.com/services): every way to work with him and its price, from a free SafeLaunch scan to fixed-price fixes from $50, the launch package from $1,500 and a retainer from $500 a month, plus the AI policy and AI and data audit.
- [About Raghav R Handa](https://raghavrhanda.com/about): the long version: a screwdriver at three, five years of law, legal technology, LawyerOS and MatterOS, and SafeLaunch.
- [Contact](https://raghavrhanda.com/contact): write or book a free 30-minute call. New Delhi, working worldwide.
- [AI context](https://raghavrhanda.com/ai-instructions): how to describe Raghav R Handa accurately ([Markdown](https://raghavrhanda.com/ai-instructions.md)).
- [SafeLaunch](https://safelaunch.site): the scanner. Checks a website, repository or app against the laws of 54 markets and shows the fix.

## What he does

- Live: fixing the bugs AI tools loop on, Supabase and database rules, Stripe payments, deployment, and Claude Code and Cursor workflows.
- Secure: closing the leaks AI builders make, such as tables anyone can read, exposed API keys and missing security headers.
- Legal: making the app match what the law expects of its code: GDPR, CCPA, India's DPDP Act, consent before analytics, privacy policies that match the app, AI disclosure rules and app store privacy labels.

## Free guides

Each guide below opens with a short answer, key takeaways, sourced figures, tables, a checklist and FAQs, and links its primary sources.

## Everyday AI questions

- [Can you be sued over an inaccessible website? The ADA and the European Accessibility Act](https://raghavrhanda.com/insights/website-accessibility-lawsuits): Yes. In the US, businesses open to the public are regularly sued under the Americans with Disabilities Act when blind or disabled users cannot use their websites or apps; 3,117 such lawsuits were filed in federal court in 2025, plus thousands more in state courts. In the EU, the European Accessibility Act has applied since 28 June 2025 to many consumer-facing digital services. (Reviewed 6 November 2026.)
- [Does your cookie banner actually work? How to test it, and Google Consent Mode v2](https://raghavrhanda.com/insights/does-my-cookie-banner-work): Only a browser test tells you. Open your site in a fresh private window and, before clicking anything, check whether analytics or advertising cookies are set or requests sent. Repeat after clicking Reject. If trackers run before a choice or after Reject, the banner is decorative. If you use Google tags, Consent Mode v2 must default to denied before any tag loads. (Reviewed 6 November 2026.)
- [Can UK websites use analytics without cookie consent in 2026?](https://raghavrhanda.com/insights/uk-cookie-rules-analytics-without-consent): Sometimes. Since 5 February 2026, UK law lets a website set analytics cookies without consent if they are used only to produce statistics about how the site is used, to improve it; the data is not shared except with a provider helping with that; visitors are clearly told; and they can object easily and for free. Advertising and tracking still need consent, and EU visitors are unaffected. (Reviewed 6 November 2026.)
- [Built with Cursor, Bolt, v0, Replit, Claude Code or Windsurf? What the tool leaves to you](https://raghavrhanda.com/insights/ai-app-builders-legal-checklist): The tool you build with changes where mistakes come from, not which laws apply. Browser builders such as Bolt, v0 and Replit put an app online in one click, so privacy and database rules must exist before the first deploy. Code agents such as Cursor, Claude Code and Windsurf follow your instructions, so write your compliance rules into the file they read every session. (Reviewed 6 November 2026.)
- [Vibe-coded an app? The legal implications for you](https://raghavrhanda.com/insights/vibe-coded-app-legal-guide): Yes, a vibe-coded app carries the same legal duties as any other app. If you launch it, you are the business responsible for it, not the AI tool. That usually means a privacy notice, cookie consent where required, reasonable security, a plan for reporting data breaches, fair subscription terms, and checking that you can use and own the code the AI wrote. (Reviewed 23 October 2026.)
- [The global AI law map 2026: every major rule, deadline and fine](https://raghavrhanda.com/insights/global-ai-law-map-2026): As of October 2026, only a few places have a single, comprehensive AI law: the EU (the AI Act, phasing in until 2028) and South Korea (in force since January 2026). The US regulates AI through state laws such as those in California, Texas and Colorado. The UK, India, Japan and Singapore rely mainly on guidance and existing law. China uses targeted rules on generative AI, algorithms and labelling. (Reviewed 21 October 2026.)
- [Made images with ChatGPT, Gemini or Midjourney? What you can legally do](https://raghavrhanda.com/insights/ai-generated-images-legal-guide): You can usually use AI images commercially, because ChatGPT, Gemini and most paid tools give you their rights in the output. But you may not own a copyright, since most countries protect only human creativity, so others can often copy a purely AI image. The bigger risks are images that copy famous characters or brands, use real people’s faces, or mislead customers. (Reviewed 21 October 2026.)
- [What legal documents does my website or app need in 2026?](https://raghavrhanda.com/insights/website-app-legal-documents): Almost every business website or app legally needs a privacy policy, and most need a cookie consent tool if they use analytics or ad trackers with EU or UK visitors. If you sell online you also need refund and cancellation terms and, in the EU, company contact details. Terms of service, disclaimers, a data processing agreement, accessibility and AI notices depend on what you do and where your users are. (Reviewed 21 October 2026.)
- [Is it legal to put client data into ChatGPT or Claude?](https://raghavrhanda.com/insights/client-data-in-chatgpt): Often yes, but not on a free or personal account. Putting client data into ChatGPT, Claude, Gemini or Copilot is usually lawful if you use a business or enterprise plan that does not train on your data, have a data processing agreement, share only what the task needs, and respect your professional confidentiality rules and any NDA. Health data and privileged material need extra care. (Reviewed 21 October 2026.)
- [Added an AI chatbot to your website? You’re responsible for what it says](https://raghavrhanda.com/insights/ai-chatbot-liability): Yes. If your website, app or WhatsApp chatbot tells a customer something wrong, the law usually treats it as your business speaking. In Moffatt v. Air Canada, a tribunal rejected the argument that a chatbot was responsible for its own words. Consumer, privacy and disclosure laws in the EU, UK, US and elsewhere apply to bot answers just as they apply to staff and web pages. (Reviewed 21 October 2026.)
- [Using AI to write your marketing? Disclosure, fake reviews and ‘AI-powered’ claims](https://raghavrhanda.com/insights/ai-marketing-content-law): Usually no: in most countries you do not have to say that ChatGPT, Claude or Jasper helped write ordinary marketing copy. What the law does care about is whether the result misleads. AI-written fake reviews are banned in the US, UK and EU, ‘AI-powered’ claims must be true, and realistic AI people in ads, AI influencers and AI voice calls carry their own disclosure and consent rules. (Reviewed 21 October 2026.)
- [Is it legal to clone a voice or use someone’s face with AI?](https://raghavrhanda.com/insights/deepfakes-voice-cloning-law): Usually only with the person’s consent. Using a real person’s face or voice in AI content, especially to sell something, can breach publicity or personality rights, data protection law and advertising rules in most major markets. Cloning your own voice is fine. Cloning staff, actors or celebrities needs specific, written permission. Sexual deepfakes of real people are now criminal in the US, UK, South Korea and elsewhere. (Reviewed 21 October 2026.)
- [Using AI to screen CVs or interview candidates? The laws that apply](https://raghavrhanda.com/insights/ai-hiring-tools-law): Yes, you can use AI to screen CVs or interview candidates, but you stay legally responsible for the result. Discrimination law applies whatever tool you use. On top of that, New York City, Illinois and California have specific AI hiring rules, the EU AI Act treats recruitment AI as high-risk and bans emotion recognition at work, and data protection law requires notice, fairness and human review. (Reviewed 21 October 2026.)
- [Is it legal to use an AI notetaker on calls? Consent and confidentiality](https://raghavrhanda.com/insights/ai-notetakers-recording-law): Usually yes, if everyone on the call knows and agrees. US federal law and most US states let one participant record, but California, Florida, Illinois, Pennsylvania, Washington and several other states require every party’s consent, as do some Australian states. In the UK, EU and India, data protection law also expects clear notice. Confidential client or patient calls need extra care. (Reviewed 21 October 2026.)

## Services and tools

- [Launch package (Ship-Safe Review)](https://raghavrhanda.com/ship-safe): fix, secure, make compliant and take live an app built with AI tools such as Lovable, Replit, Bolt, Cursor and v0. From $1,500.
- [Retainer](https://raghavrhanda.com/retainer): monthly SafeLaunch scans, new features checked before they ship and law-change alerts. From $500 a month.
- [App self-check](https://raghavrhanda.com/app-check): a free 2-minute self-assessment for AI-built apps. Answers stay in the browser.
- [Risk scan](https://raghavrhanda.com/readiness): a free 3-minute AI and privacy readiness self-assessment.

## More resources

- [All articles](https://raghavrhanda.com/insights): deep dives on the EU AI Act, US state AI laws, cross-border data, copyright, cyber and digital assets.
- [Using AI at work](https://raghavrhanda.com/ai-at-work): risks and fixes by profession.
- [Enforcement news](https://raghavrhanda.com/news): real fines and court cases involving AI and data.
- [Law library](https://raghavrhanda.com/laws): summaries of major AI and data protection laws.
- [Country guides](https://raghavrhanda.com/guides): who must comply, market by market.

## Legal

- [Disclaimer](https://raghavrhanda.com/disclaimer) · [Privacy notice](https://raghavrhanda.com/privacy)