New “Improvement Notice” Concept To Replace Penalty For First Offence In Jan Vishwas 2.0 Bill: Report

The Centre will table the Jan Vishwas (Amendment of Provisions) Bill, 2025, in the Lok Sabha on Monday (18 August), which will introduce a new concept of “improvement notice” instead of a penalty for first-time offences, reported The Indian Express.

The Union Cabinet cleared the Bill last week, aiming to amend multiple enactments, decriminalise provisions, and promote trust-based governance for ease of living and doing business.

Finance Minister Nirmala Sitharaman had announced the move in her 1 February 2025 Budget speech.

She said, “In the Jan Vishwas Act 2023, more than 180 legal provisions were decriminalised. Our government will now bring up the Jan Vishwas Bill 2.0 to decriminalise more than 100 provisions in various laws.”

The proposed Bill shifts from the “penalise on first detection” model of Jan Vishwas 1.0 to an “inform–correct–penalise” framework.

It defines the “improvement notice” mechanism and applies it across multiple sections.

First-time offenders will receive a notice and time to correct violations, while penalties will apply from the second offence onwards.

The fine for repeated offences will remain the same as in Jan Vishwas 1.0, with higher penalties for subsequent violations, capped under respective sections.

Officials say this structure ensures full decriminalisation while encouraging voluntary compliance and reducing fear of punitive action for minor, unintentional mistakes.

In his Independence Day speech, Prime Minister Narendra Modi said, “In our country, there are such laws that can put people in jail for very small things — you would be shocked… We had introduced a Bill in Parliament earlier, and we have brought it again this time.”

Earlier, the Jan Vishwas Act, 2023, had scrapped imprisonment provisions under several laws, including the Food Corporations Act, 1964 and the Indian Forest Act, 1927.

Also Read: Nirmala Sitharaman Tables Revised Income Tax Bill 2025 In Lok Sabha, Accepts Most Panel Suggestions