Karnataka Rohit Vemula (Prevention of Discrimination, Exclusion or Injustice) (Right to Education and Dignity) Bill, 2026get Cabinet approval in April After months of discussion and debate.

The amended Bill aims to prevent exclusion and injustice while protecting the right to education and dignity of students, teaching faculty and non-teaching staff belonging to Scheduled Castes and Scheduled Tribes (SC/ST). It guarantees equal access to education in all higher educational institutions across Karnataka, including public, private, charitable institutions and deemed universities.

Brief history of the bill

am following Rohit Vemula’s deathA 26-year-old Dalit PhD scholar at the University of Hyderabad, in 2016, political and civil society pressure is building for a dedicated anti-caste discrimination law in higher education.

The Karnataka government started drafting the bill almost a year ago in April 2025, after Rahul Gandhi wrote to the state urging it to introduce the Rohith Vemula Act.

The idea was to create a framework encompassing preventive and remedial measures to address the issues of students and teachers from marginalized communities who continue to face caste discrimination on campuses.

The draft bill went through several iterations. During the Karnataka Budget 2026–27, Chief Minister Siddaramaiah formally announced the government’s intention to introduce the Rohith Vemula Bill. This bill got approval from the state cabinet in April.

What is defined in the Bill?

The Bill defines both caste-based harassment and discrimination, including direct and indirect discrimination.

Caste-based harassment is defined as any hostile, derogatory or offensive behaviour, whether verbal, non-verbal, oral, written, recorded, visual or digital, that humiliates or threatens the dignity of SC/ST persons on the basis of caste.

Discrimination is any intentional or unintentional act or omission that adversely impacts people from SC/ST communities, including failure to treat them equally as others, preferential treatment towards castes considered ‘higher’ in the caste hierarchy, and other similar caste-based practices.

Who is the victim? Do the provisions apply only inside classrooms?

The aggrieved person may be a student or faculty member or employee of the SC/ST community in any higher educational institution who has allegedly faced discrimination and/or caste-based harassment.

The Bill covers not only the places of work and residence within educational institutions, but also all online and offline spaces where students, faculty and staff socialize with each other. These include hostels, health centres, sports grounds, staff quarters, parks, roads, canteens and transport facilities.

What are some important provisions of the bill?

The Bill states that institutions of higher education shall be open to all persons regardless of their sex, gender, race, caste, religion, sexuality or disability, and clearly states that no person belonging to SC/ST communities should be subjected to discrimination or caste-based harassment.

It says, “It shall not be lawful for any institution of higher education to impose on any person any test of religious belief or profession in order to entitle him to be admitted thereto as a teacher or student or to hold any office thereunder or to graduate therefrom or to enjoy or exercise any privilege prescribed.”

The Bill recognizes the right of individuals from SC/ST communities to organize collectively, form unions, promote anti-caste discussions through reading clubs, refuse to participate in programs and activities that reinforce caste prejudice, and not be subjected to punishment or negative assessment for asserting their views against caste. It also supports the right of community members to maintain privacy regarding their caste identity.

What can be called caste-based oppression?

One of the most detailed sections of the bill is the list of prohibited acts. It identifies 30 prohibited discriminatory acts on the basis of race.

These include admissions and hiring practices, such as asking questions that probe social background or family occupation in a discriminatory or disrespectful manner; Denial of admission despite meeting admission criteria; Accepting or rejecting applications on the basis of caste; And violation of reservation policies.

It also includes discrimination within educational spaces, including segregation within classrooms or educational settings; exclusion from class, laboratory, or extracurricular activities; unequal access to faculty contacts; Discriminatory assessment in examinations; And forcing students to do activities they would not normally do. The Bill also bans curriculum content that glorifies certain castes or denigrates SC/ST communities.

Institutional and administrative barriers are also addressed, such as ensuring access to opportunities that favor non-SC/ST individuals; Arbitrarily forcing students to enroll in particular courses; changing norms to their detriment; excluding them from committees or decision-making bodies; and failed to provide basic student facilities.

Financial exploitation is another category, which includes demands for payments beyond declared admission policies, forced financial contributions, withholding of documents for money and arbitrary delays in scholarship distribution.

The bill prohibits coercive practices such as forcing people to participate in religious or cultural events against a student’s beliefs, taking on uncompensated labor beyond specified duties, and steering individuals into inappropriate occupations or career paths based on race.

It addresses stigma and restriction of rights, including disclosing admission status through reservation categories, labeling students as “reserved category” or “quota students”, attributing academic poor performance to caste, barring participation in democratic or secular associations, restricting reporting of unfair practices, and limiting access to internships or work opportunities of their choice.

Not taking action on a discrimination complaint within the time prescribed under the Act and creating an intimidating or hostile environment for the aggrieved person after filing the complaint would also amount to caste-based harassment.

What is the prevention mechanism?

According to the Bill, every higher educational institution will have to constitute an Equity Committee, in which at least 80% of the members will be from the SC/ST community. The committee should consist of a Presiding Officer who is from the SC/ST community, and above the rank of Assistant Professor in the institution, two staff/faculty members, one administrative staff, one member of the NGO committed to the cause and two students from the SC/ST community, one of whom should be a woman.

The aggrieved person can submit a complaint to the presiding committee within three months from the date of the incident. If the complainant does not agree to conciliation, the Committee will have to initiate an investigation into the complaint.

If Prima facie If any case of atrocities comes to light, a complaint has to be lodged with the law enforcement agencies within seven days. The committee may recommend suspension of the accused pending investigation.

What are the remedies for the victim and punishment for the perpetrators?

Measures include Administrative intervention that will eliminate the adverse impact, reasonable adjustments, monetary compensation, counseling assistance and any other measures deemed necessary.

Students found to have engaged in discrimination or caste-based violations may face penalties, including exclusion from hostels, campus, examinations and other activities; withholding results; suspension; Or expulsion.

Faculty or staff found responsible may face adverse comments in performance evaluations, be deprived of supervisory duties, be denied employment or contract renewal, be withheld from promotion or salary increases, demotion, pay or salary reductions, suspension, or dismissal.

The severity of the penalty will be determined by factors, including the nature and severity of the harassment; the age or condition of the offender; the frequency and prevalence of the conduct; any resulting physical or mental harm; The suffering caused to the victim; and the extent of its impact on their work or academic performance.

For higher educational institutions found violating the provisions, the Bill proposes a maximum fine of Rs 10 lakh or denial of financial aid.

What you need to know about Karnataka’s Rohith Vemula Bill

The Karnataka Rohith Vemula (Prevention of Discrimination, Exclusion or Injustice) (Right to Education and Dignity) Bill, 2026 received cabinet approval in April after months of deliberations and debate. The amended Bill aims to prevent exclusion and injustice while protecting the right to education and dignity of students, teaching faculty and non-teaching staff belonging to Scheduled Castes and Scheduled Tribes (SC/ST). It guarantees equal access to education in all higher educational institutions across Karnataka, including public, private, charitable institutions and deemed universities. | Video Credit: The Hindu

Often, critical information related to policy making to ease the lives of citizens exists in government databases, regulatory portals, certification systems and official statistical repositories, which are accurate, structured and authoritative, yet largely inaccessible to analytical and modern AI tools. While AI systems can generate fluent answers from scraped Internet content, they remain isolated from government databases by their design.

As a result, when citizens, policy-makers, or entrepreneurs seek guidance on questions related to regulatory and planning requirements, such as government certification or eligibility for government schemes, using AI tools, AI produces informed guesses rather than definitive answers, not because the knowledge is absent, but because it cannot access the systems where that knowledge actually resides.

As citizens and enterprises alike increasingly turn to AI to unearth such information, and its use is expected to increase as awareness of AI’s capabilities increases, the need for AI-ready data is more urgent than ever.

India generates huge amounts of data in agriculture, commerce, health, education and other civil services.

The e-statistics portal of the National Statistical Office alone contains 135 million records of various socio-economic indicators such as GDP, industrial production, consumer prices and labor force surveys. Through its Digital India initiative, the government has already taken important steps in digitizing public services and improving access to schemes and regulatory information through UMANG, MyScheme and several other portals, generating information on a large scale.

Although these efforts have made information more accessible to citizens and businesses, data remains dispersed across multiple platforms and departments. The next step is to extend this remarkable digital transformation to AI systems by making our information portals interoperable with them, enabling a simple natural-language query to get accurate, authoritative answers directly from these government sources and providing accurate guidance to citizens and businesses in real-time.

Data exists in a number of distinct systems and portals, each designed to serve specific purposes and audiences. A wealth of resources spanning departments, documents and datasets, each providing unique insights, although often in different formats and contexts, is available; There is a need to make them AI-ready. Linking these resources effectively can empower citizens and businesses with a more holistic and actionable view of the information available. AI has already proven it can operate at this scale, processing huge datasets, converting unstructured documents into usable formats, and answering complex questions in natural language.

The opportunity lies in applying this technology to government data. By bridging the gap between AI tools and the rich information already available, we can unlock economic value for millions of users, including enterprises.

Reducing information friction benefits citizens and businesses alike, but can be especially transformative for small businesses and the marginalized, for whom access to timely intelligence and regulatory guidance can open up opportunities that were previously difficult to access. One of the best use cases could be in the MSME sector.

When these enterprises can efficiently access market information, compliance requirements and plan eligibility, they compete more effectively in global markets, optimize their operations and grow faster. Manufacturing competitiveness strengthens, employment expands in cities and towns, and supply chains become more flexible. The multiplier effect of improved information access on this huge enterprise base translates into significant economic acceleration. With such enablement, the MSME sector can truly act as a champion as envisaged in this year’s Budget.

All this is possible only when we have AI-ready data. A weak data foundation cannot be compensated for by AI. NSO India, being the nodal agency for official statistics for the Government of India, has taken several steps for data harmony such as national meta data structure, statistical quality assurance framework, compilation of unique identifiers, and codes and taxonomies to be used while creating planned data layers for interoperability.

The data accessibility requirement is met through a digital bouquet for data dissemination consisting of a website, eSankhyiki portal, mobile app, microdata portal and metadata portal. These applications provide government data in an interoperable manner using APIs for consumption by both humans and machines. Based on these parameters, NSO India has AI-ready data for use.

On February 6, 2026, the National Statistical Office, India took a significant step forward by launching a Model Context Protocol server, which maintains an open technology layer on the eSankhyiki repository. This infrastructure allows AI systems to connect directly to datasets on eSankhyiki and query them programmatically. Now anyone can ask an AI assistant about pricing trends, employment patterns or industrial production and get answers synthesized from official data in real time.

Market analysis shows that previously sought-after specialized expertise is becoming accessible through natural language queries. Immediate use cases started coming from users, from dashboard creation to deep-dive data analysis. This represents a fundamental shift in how government data serves the economy.

In an AI-driven world, more than just publishing and organizing data is required. It must be reliable, interoperable, relevant and machine-usable. Users need to know where the data comes from and whether they can trust it. Privacy protection needs to be built into the architecture, not added as an afterthought. The governance framework needs to clearly define lawful use. Quality assurance needs to be maintained continuously, flagging anomalies before they are propagated through the AI ​​system.

These elements determine whether AI produces reliable outputs and builds public trust. These safeguards are included by design in the NSO implementation. MCP provides access to verified official data through server controlled protocols while keeping the data close to its official source. When datasets are standardized, agreed upon, and made searchable through common APIs, they become an AI-ready foundation on which any model or application can safely build.

The initial deployment includes seven datasets: Periodic Labor Force Survey, Consumer Price Index, Annual Survey of Industries, Index of Industrial Production, National Accounts Statistics, Wholesale Price Index and Environmental Statistics, with more to come in the future. Each may be interrogated individually or in combination.

Consider what this makes possible. An MSME exploring new markets can analyze employment trends and industrial potential in the target sectors through an interaction with the AI ​​assistant. A researcher studying regional economic patterns can combine industrial production data with labor force surveys without manually reconciling different formats and time periods. Infrastructure development occurs through progressive integration. As more authoritative data sources join the network, queries can integrate them with other datasets. Each new connection expands and multiplies the analytical possibilities across all existing sources, enabling deeper insights and more comprehensive decision making.

MOSPI’s data reconciliation and dataset linking initiatives are laying the foundation for a data ready future. Government datasets can become accessible foundations upon which any developer, researcher, or enterprise can build. Small businesses gain access to market information that was previously only available to large corporations. Policy makers receive real-time feedback. Researchers connect datasets that were previously incompatible.

As datasets connect across the network, each connection multiplies the analytical possibilities across all existing sources. Economic benefits will increase as the number of adoptions increases. When data becomes a horizontal capability that spans all sectors, innovations stop being trapped in silos.

This approach recognizes a fundamental principle: those who generate the data should benefit from it. India’s competitive advantage in the AI ​​era will come from responsibly mobilizing diverse, relevant, high-quality data.

The question is no longer whether data can create economic value. The question is whether we design systems where that value is shared widely to empower both prosperity and the public good. AI-ready data infrastructure, built with a foundation of trust and security measures, transforms scattered information into accessible intelligence.

Thus India builds the data foundation that powers AI for economic growth at population scale en route to a developing India.

(Views expressed are personal)

This article is written by Saurabh Garg. Secretary, Ministry of Statistics and Program Implementation and Shalini Kapoor, Chief Strategist, Data and AI, Onestep Foundation.