Genera ve Ar ficial Intelligence in Educa on and Research: User Guidelines and Policy
1. Preamble
At NMIMS, we stand at the forefront of technological advancement, embracing the transforma ve poten al of Genera ve AI (Gen AI) to revolu onize our educa onal landscape. We recognize that Gen AI holds immense promise in enhancing crea vity, fostering innova on, anddrivingproduc vityacrossamyriadofacademicandadministra ve domains. As we integrate Gen AI into our university's fabric, we do so with a clear vision of harnessing its power to enrich educa onal experience, advance research capabili es, and streamline administra veprocesses.
Our posi ve stance on Gen AI is underscored by our commitment to prepare our students and faculty for a future where ar ficial intelligence plays a pivotal role in various sectors. We believe that by incorpora ng Gen AI into our curriculum, research, and opera ons, we can cul vate a genera on of forward-thinking leaders who are adeptat leveragingcu ng-edgetechnologiestoaddresscomplexchallenges.
However, we recognize that with the great poten al of Gen AI comes theresponsibilityofensuringthatitisusedethicallyandresponsibly.At NMIMS, we are keenly aware of the need for robust checks and balances to govern the use of Gen AI. We understand that different typesofuserswithintheuniversity—students,faculty,researchers,and administrators—have varied needs and responsibili es. Therefore, our guidelines are designed to provide clear and specific direc ves for each usergroup,ensuringthatGenAIisu lizedinamannerthatupholdsour valuesofethicalstandards,academicintegrityanddataprivacy.
For students, our guidelines emphasize the importance of using Gen AI as a tool for learning and innova on while maintaining the originality and authen city of their academic work. For faculty and researchers, the focus is on harnessing Gen AI to push the boundaries of knowledge and innova on, while adhering to ethical research prac ces. For administra vestaff,GenAIisseenasameanstoenhanceefficiencyand service delivery, with strict protocols to protect sensi ve informa on andmaintaindatasecurity.
Through though ul integra on and mindful oversight, NMIMS University is dedicated to leveraging Gen AI to create a brighter, more compassionate future. We are commi ed to fostering an environment where Gen AI is a catalyst for posi ve change, driving us towards our goal of becoming a leader in technological and educa onal excellence. By se ng clear guidelines and fostering a culture of responsible AI use, we aim to ensure that Gen AI serves as a powerful ally in our mission to educate,innovate,andinspire.
2. AboutthePolicy
2.1.Scope:ThispolicyappliestoallmembersoftheNMIMScommunity, includingstudents,faculty,staff,andresearchers.
2.2.Ra onale: Gen AI tools have a far-reaching impact on teaching, learning and research. As such, a need was felt to make the University's stance clear, to offer guidelines on responsible usage and to make known the consequences of misuse of Gen AI tools. The University recognizes the poten al of Gen AI yet is mindful of its ill effects and thereforeadoptsaflexibleapproachtowardsGenAIusage–onethaton
a case-by-case basis may range from whole-heartedly embracing the technologytousingitwithcau onanda ribu ontoprohibi on.
2.3.Feature: Given the fast-paced and unpredictable changes in Gen AI'scapabili esanditsimpactoneduca onandresearch,thepolicywill berevisitedandrevisedasandwhenrequired.
3. EthicalUse
3.1.Respect for human dignity: Respect for human dignity is understood as a cornerstone principle and one that is non-nego able. The principle emphasizes the ethical responsibility we all share when u lizing these tools. Here's a breakdown of what it means in this context:
3.1.1. Upholding Fundamental Rights: Genera ve AI should never be used to infringe upon the fundamental rights of individuals. This includes, but is not limited to, the right to privacy, freedom from discrimina on,andprotec onfromharm.
3.1.2. Prohibi on on Offensive Content: The crea on and dissemina on of content that is offensive, hateful, or discriminatory is strictly prohibited. This includes content that targets individuals or groups based on religion, gender, caste, race, sexual orienta on, disability,oranyotherprotectedcharacteris c.
3.1.3. Protec ngPrivacy:Whenusinggenera veAItools,usersshould be mindful of the data they input. Users are warned against inpu ng privateorconfiden alinforma on.Addi onally,theoutputsgenerated byAItoolsshouldnotrevealsensi vedetailsaboutindividuals.
3.1.4. Promo ng Respec ul Communica on: Genera ve AI can be a powerful tool for communica on. However, it is crucial to ensure that the content generated fosters respec ul and inclusive dialogue. Gen AI should not be to create content that dehumanizes anyone, promotes stereotypes,orincitesviolence.
3.2.AcademicIntegrity:NMIMSisaforward-lookinguniversity.Itisalso steadfastly commi ed to nurturing a culture of academic honesty and integrity.Assuch,weembraceandencouragetheuseofGenAItoolsfor the revolu onary poten al they hold. Simultaneously though we are cau ous regarding the misuse of these tools in ways that hinder true learning, skill development, originality and what in summary we might call the development of human intelligence. In principle, the University encourages a flexible framework where the nature of the task determines the correct, ethical use of these tools. The subsequent sec onsofthepolicyprovideclearguidelinesregardingresponsibleuse andtheconsequencesofthefailuretodoso.
4. Confiden alityandDataPrivacy
NMIMS steadfastly upholds the principle of protec ng the confiden ality of all stakeholders and the University. All users of Gen AI toolsmustthereforeeducatethemselvesaboutthepoten althreatsto confiden ality that may arise due to such usage. They must be able to: iden fy confiden al data, be cau ous while inpu ng any data, be cau ous while dissemina ng any output generated by Gen AI, familiarize themselves with data protec on laws and seek guidance whenindoubt.
4.1.Iden fying Confiden al Data: Recognize the types of data considered confiden al within NMIMS. This might include personally iden fiable informa on (PII) of students, faculty, staff, or researchers. Addi onally, proprietary informa on, trade secrets, and intellectual propertyfallundertheumbrellaofconfiden aldata.
4.2.Securing Confiden al/Personal Data: Users must refrain from feeding any form of confiden al or sensi ve data into genera ve AI systems. This includes names, addresses, phone numbers, financial informa on, or any data that could be used to iden fy or harm individualsorcompromiseprivacyorintellectualproperty.
4.3.Data Protec on Laws: Users should familiarize themselves with lawssuchasIndiaDigitalPersonalDataProtec onAct2023(DPDPA),EU GDPR (General Data Protec on Regula on) or US HIPAA (Health Insurance Portability and Accountability Act) depending on the type of data involved. These laws outline specific requirements for data collec on,storage,anduse.
4.4.Understanding AI Tool's Data Prac ces: Users should inves gate the data prac ces of the genera ve AI tools they intend to use. Some tools may anonymize user data, while others may retain it for improvement purposes. Understanding these prac ces is crucial for ensuringconfiden ality.
4.5.Seeking Clarifica on: If users are unsure about the confiden ality status of data or have ques ons regarding its use with genera ve AI, theymustconsulttheNMIMSITdepartment.
5. ResponsibleDeployment
5.1.Informed Use: While Gen AI can offer content at the click of a bu on, users need to be aware that this is unverified content. The followingmeasuresensuremoreresponsibleuseofthistechnology:
5.1.1. UnderstandingLimita ons:
5.1.1.1. Awareness of Algorithmic Biases: Genera ve AI tools are trained on massive datasets. However, the tools are suscep ble to biases present within the data they are trained on. For instance, a language model trained on biased news ar cles may perpetuate stereotypes in its generated text. These biases can manifest in factual inaccuracies, skewed perspec ves, or unfair representa ons in the generated content. Users can mi gate bias by selec ng a tool wisely. Users must look for tools that have been trained on diverse datasets to minimizetheriskofbiasinthegeneratedcontent.
5.1.1.2. Disclosure of Limita ons: When incorpora ng AI-generated content into one's work, users must be aware of the limita ons of the toolandcontentgenerated.Theyshouldacknowledgethepossibilityof biasandtheneedforhumanverifica on.
5.1.2. Verifica on before Use: Users should not blindly accept AIgeneratedcontentasfactualandcorrect.Theyshouldcri callyevaluate theoutputsforaccuracyandreliability.Therearevariouswaysinwhich one can verify content, some of these include: using fact checking tools such as Snopes, SciCheck, Media Bias/Fact Check; seeking expert review; seeking reviews from colleagues or reviewers; referring to publishedresearch.
5.1.3. Awareness of and Adherence to User Agreements: Providers of Gen AI tools state terms and condi ons outlining how user data is collected, stored, and used. They may specify limita ons on data types, anonymiza on requirements, or restric ons on commercial use of generated content. Users must carefully review these terms before accep ng them. Once accepted, following these terms ensures responsible data usage prac ces and avoids poten al legal or ethical disputes.
5.1.4. Competent Use: Users should cul vate a cri cal approach while evalua ng the informa on generated by Gen AI. This includes ques oning the source, purpose, and poten al biases inherent in AIgenerated content. They should be transparent and demarcate the contribu on of AI content as well as discuss the strengths and limita ons caused by using AI tools. They should at all mes consider theethicalimplica onsoftheiruseofGenAI.
5.1.5. A ribu ons: Users must clearly iden fy AI generated content and use different forma ng styles for human and AI content. Correctly ci ng AI content involves including the name of the AI tool, the date of access,andanyrelevantparametersusedforgenera on.
5.2.Adherence to Non-Malicious Use: NMIMS is commi ed to fosteringacommunitybuiltontruth,integrity,andethicalconduct.This extends to the use of genera ve AI tools, which hold immense power but also carry the poten al for misuse. As such, all stakeholders at NMIMS must ensure that the technology is not used for any harmful purposes such as, but not limited to - fabrica ng news, or data, or records; spreading falsehood in any form; crea ng, retrieving or
circula ng informa on that could cause harm. AI should also not be used for malicious purposes such as iden ty the , phishing scams, or financialfraud.
6. TrainingandSupport
6.1.Training: NMIMS recognizes the importance of con nually educa ng its community towards: advancements in the field of Gen AI, its poten al, its risks, its limita ons, and methods of fair use. The university also recognizes the need to impart technical training to its stakeholders so that users may make the most of Gen AI tools. Towards thisend,theuniversitywillhostworkshops,guestlectures,andtraining sessionsandshareinforma onvianewsle ers,blogsandemails.
6.2.Support:NMIMSrecognizesthatnaviga ngthecomplexi esofGen AI can be challenging. To empower its community towards responsible AIuse,NMIMSoffersacomprehensivesupportsystem.
6.2.1. Ethics Advisory Groups: The University's ethics advisory groups having experts from computer science, AI, social science, law, philosophy, ethics and school/domain specific experts will guide researchers and students in ma ers such as fair use, risks, and a ribu on.
6.2.2. Technical Support Teams: These teams will provide users with technical assistance, such as: troubleshoo ng technical problems encountered while using genera ve AI tools; answering ques ons about the technical capabili es and limita ons of specific tools; providing guidance on op mizing use of genera ve AI tools for desired outcomes.
6.2.3. Centralized Online Repository: An online repository comprising FAQ's, tutorials, case studies, training modules will be available for con nuouslearning.
6.3.BenefitsofU lizingSupportResources:
6.3.1. Promo ng Confidence: Access to expert guidance and technical support fosters confidence in naviga ng the complexi es of genera ve AI.Thisempowersuserstoleverageitspoten alwhilemi ga ngrisks.
6.3.2. Ensuring Compliance: By consul ng with support resources, users can ensure that the use of genera ve AI adheres to NMIMS policies,dataprotec onlaws,andethicalbestprac ces.
6.3.3. Fostering Collabora on: The support system encourages collabora on within the NMIMS community. Users can learn from the experiencesofothersandatthesame mecontributetheirknowledge topromoteresponsibleAIuseacrossdisciplines.
6.4.NMIMS's Commitment to Ongoing Support: The University is commi edtocon nuallyrefiningitssupportsystem.Thisincludes:
6.4.1. User Feedback Mechanisms: NMIMS will establish channels for userstoprovidefeedbackontheeffec venessofthesupportresources andiden fyareasforimprovement.
6.4.2. Expanding Exper se: The University will con nually seek the guidance of experts in the field to ensure adequate and mely response.
7. Compliance-MonitoringandRepor ng:
NMIMS is commi ed to fostering a culture of responsible Gen AI use within its community. This sec on of the Genera ve AI User Guidelines and Policy outlines the monitoring and repor ng mechanisms implementedtoachievethisgoal.
7.1.Monitoring:NMIMSreservestherighttomonitortheuseofGenAI tools within the university. This monitoring is not intended to be intrusive but rather serves to ensure that users are adhering to the principlesoutlinedinthispolicy.
7.1.1. Purpose of Monitoring: Monitoring may be carried out for purposes such as: iden fying poten al misuse of genera ve AI toolssuch as crea ng false informa on or engaging in malicious ac vi es; verifying compliance with data privacy regula ons; assessing the overalleffec venessofthepolicyinpromo ngresponsibleAIuse.
7.1.2. TransparencyinMonitoring:
• Respect for Privacy: NMIMS will priori ze user privacy while conduc ng any monitoring ac vi es. The university will only monitor data relevant to ensuring adherence to this policy and will anonymize datawheneverpossible.
• Communica on of Monitoring Prac ces: NMIMS will keep the university community informed about its general monitoring prac ces relatedtogenera veAIuse.
7.2.Repor ng: Repor ng is a shared responsibility of all NMIMS stakeholders. If one suspects that someone is misusing genera ve AI
tools, it is crucial to report it to the designated university authori es. There are several benefits of being alert and repor ng any misuse of Gen AI. Some of these benefits are: addressing a poten al misuse of genera ve AI before it escalates into a serious issue;and crea ng a safe environmentthatupholdsintegrity,privacy,anddignity.
TheUniversitywill:
7.2.1. Provide Confiden ality and Support: The University will ensure confiden ality for those repor ng misuse and will provide support throughouttheinves ga onprocess.
7.2.2. Create Opportuni es for Open Communica on: The University will organize forums and discussions where the community can discuss concerns, share best prac ces, and collec vely contribute to a culture of responsible AI use. In these forums, users can also discuss the university'sGenAIpolicyandprovidefeedbacktoensurethatthepolicy iscomprehensive,prac cal,andadaptabletotheevolvinglandscapeof genera veAI.
8. Enforcement
The preceding sec ons have provided clear guidelines regarding the ethical, responsible deployment and integra on of Gen AI tools. This sec ondetailsthepoten alconsequencesofviola ngtheguidelines.
8.1.Spectrum of Viola ons: The severity of consequences will depend on the nature and intent of the policy viola on. Minor infrac ons, like uninten onal use of biased data, might be addressed through educa onal interven ons. More serious viola ons such as deliberately crea ngharmfulcontentcouldleadtostricterrepercussions.
8.2.DisciplinaryAc ons:Poten alconsequencesforviola onsinclude:
8.2.1. SuspensionofAIToolAccess:Individualswhomisusegenera ve AI tools may have their access suspended for a specific period. This allows meforreflec onandretrainingselfonresponsibleAIuse.
8.2.2. Academic Penal es: For academic misconduct involving genera ve AI, penal es may range from deduc on of marks to failing grades or, in extreme cases, expulsion from academic programs dependingontheseverityoftheviola onanduniversityregula ons.
8.3.EmphasisonEduca on:
8.3.1. Correc ve Measures: The primary focus of consequences is not puni ve but rather correc ve. NMIMS aims to educate individuals about the ethical implica ons of their ac ons and provide opportuni es to learn from their mistakes. This will involve mandatory training sessions on responsible AI use for both faculty and administra vestaff.
8.3.2. Promo ng a Responsible Culture: By addressing misuse construc vely, NMIMS aims to create a culture where responsible AI use is the norm. Open communica on and learning from past incidents arecrucialaspectsofthisendeavour.
8.4.Con nuousRevisions:
8.4.1. Reviewing Consequences: NMIMS will periodically review the established consequences for policy viola ons to ensure they are effec veandpropor onatetotheseverityofthemisuse.
8.4.2. Adap ng to Change: As the field of genera ve AI evolves, NMIMSmayneedtoadaptitsapproachtoaddressingmisuse.Thismay involve addressing new types of viola ons or revising exis ng consequences.