Do I have to disclose that AI helped write a letter?

Indian law does not generally require it for ordinary text documents, as of October 2026. Your engagement terms, professional body or the recipient’s rules may, so check them.

Does India require AI images to be labelled?

Intermediaries that let users create synthetic audio or visual content must label it prominently under rule 3(3) of the IT Rules, as amended on 10 February 2026.

Does a chatbot on our website need to say it is AI?

If it serves people in the EU, Article 50 of the EU AI Act requires that, unless it is obvious from the context. Saying so is good practice everywhere.

Can an AI tool be listed as a co-author?

Not under ICMJE’s recommendations: chatbots cannot be responsible for the accuracy, integrity and originality of the work, so they cannot be authors.

ANVE: the private AI office, on the record. See what that means

ANVE

Private AI glossary · updated 2026-10-07

AI disclosure
Telling the people who rely on your work that AI helped make it, and how.

Short answer. AI disclosure is telling the people who receive or rely on something that AI was used to make it, and how. The rules differ by setting. In India, the IT Rules amendment notified on 10 February 2026 requires intermediaries whose services can create synthetic audio, images or video to label it prominently and embed provenance data; MeitY’s FAQ says pure text is not covered, and routine preparation of documents and presentations is excluded. In the EU, the AI Act requires chatbot providers to tell people they are talking to AI and deployers to disclose deepfakes; the European Commission says these transparency rules took effect in August 2026. Medical journals following ICMJE require authors to disclose AI use at submission. For ordinary office work, disclosure duties usually come from client contracts and professional rules, so check those.

What the main rules require

Rule	Who it binds	What must be disclosed
IT Rules, 2021, rule 3(3), inserted by G.S.R. 120(E) of 10 February 2026	Intermediaries offering tools that can create synthetic audio, visual or audio-visual content	A prominent label, or a prefixed audio disclosure, plus embedded metadata or provenance where technically feasible
EU AI Act, Article 50	Providers of chatbots; deployers of deepfakes and of AI-generated text published to inform the public	That the person is interacting with AI; that content is artificially generated, unless text has had human editorial review
ICMJE Recommendations	Authors submitting to journals that follow ICMJE	Whether and how AI-assisted tools were used; chatbots cannot be listed as authors

What India’s rules do not cover

The 2026 amendment defines synthetically generated information as audio, visual or audio-visual content made to appear real. The definition excludes routine or good-faith editing, and routine creation of documents, presentations, PDF files, training material and research outputs that do not create a false document or record. MeitY’s FAQ adds that pure text outputs are not synthetically generated information, though unlawful text is still caught by other rules.

So the IT Rules do not require you to label an AI-drafted board note. Whether your client, court or regulator expects disclosure is a separate question.

A practical disclosure, by setting

Client reports: a line in the engagement letter and the report, for example “Prepared with the help of AI tools; reviewed and approved by [name].”
Court filings: follow the court’s practice directions, and verify every citation by hand.
Research papers: describe the tool and its use where the journal asks, as ICMJE requires.
Phone and chat assistants: say at the start that the caller is speaking to an AI.
Images and video: label synthetic images, especially of real people.

Keeping evidence behind a disclosure

Disclosure is easier to stand behind when you can show what the AI did and who checked it. ANVE labels answers that come from cloud models, shows the source passage behind document answers, and logs every model call on the device, so an office can show which work was AI-assisted and who reviewed it. This page is general information, not legal advice.

Questions people ask

Do I have to disclose that AI helped write a letter?: Indian law does not generally require it for ordinary text documents, as of October 2026. Your engagement terms, professional body or the recipient’s rules may, so check them.
Does India require AI images to be labelled?: Intermediaries that let users create synthetic audio or visual content must label it prominently under rule 3(3) of the IT Rules, as amended on 10 February 2026.
Does a chatbot on our website need to say it is AI?: If it serves people in the EU, Article 50 of the EU AI Act requires that, unless it is obvious from the context. Saying so is good practice everywhere.
Can an AI tool be listed as a co-author?: Not under ICMJE’s recommendations: chatbots cannot be responsible for the accuracy, integrity and originality of the work, so they cannot be authors.

What is not real yet

No sealed ANVE unit exists; every image of the appliance on this site is a labelled render. The software runs today on a development machine. No customers yet, no certifications, no delivery date, and the final price is not announced.

Reservations are ₹999 in India or US$12 elsewhere, refundable in full until you confirm an order in writing (terms). Offices that cannot use cloud AI can apply to be one of ten design partners.

AI disclosureTelling the people who rely on your work that AI helped make it, and how.