Data Ethics And Privacy Full Course 2026 [FREE] | Data Ethics Tutorial For Beginners | Simplilearn

Key Takeaways

Hey everyone, welcome to this video on foundations of ethical data practices. Think about the last time you downloaded a new app. It immediately asked you for your location, your camera, maybe even your context, right? Did you ever stop to wonder why it needs all that or what happens to the data once you hit allow? In today's world, we are constantly exposed to digital applications and data has become the most valuable resources we have. But here's the thing. With great data comes great responsibility. There's a massive demand right now for people who don't just know how to analyze data, but how to do it ethically. This video is unique because we are not just going to talk about the dry rules. We are diving into real world scenarios like how Snapchat uses your facial data or why companies like T-Mo have faced massive ethical concerns over data breaches. If you want to build trust with your users and understand the legal frameworks that govern our digital world, you'll want to stay tuned. Here's our road map for this session. First, we'll break down the five key principles of data ethics. transparency, fairness, accountability, privacy and also security. Second, we'll look into the ethical challenges we face every day like biases in algorithms for example, how a resume screening tool might accidentally discriminate against a certain groups. Third, we'll discuss about informed consent. We'll explain why it's critical that users know how their data is collected and who has access to it. Finally, we'll talk about building a trustworthy data ecosystem. We'll cover everything from international laws like GDPR to how organization can conduct ethical assessment to stay secure. Whether you're a data scientist, a business leader, or just someone curious about your digital rights, this course is for you. Let's jump right in. Also just a quick information. If you're interested in advancing your career in data analytics and generative AI, the professional certificate course in data analytics and generative AI by IIT Kpur is perfect opportunity for you. This course delivered by the esteemed ENICT Academy provides not just theoretical knowledge but also practical hands-on learning to help you master industry relevant tools like Excel, SQL, Python, R, PowerBI, Tableau and cuttingedge generative AI technologies. In this program, we'll also dive deep into data ethics. A critical area where you'll learn how to handle data responsibility, ensure privacy, and make ethical decisions while working with data. Ethical data practices are essential especially as AI and data analytics continue to shape industries worldwide with live master classes from IIT Kpur faculty 12 plus projects and access to exclusive industry tools. This course is designed to give you skills needed to excel in datadriven decision-m and stay ahead in fast evolving field. So why wait? Apply now and get started on your path to a rewarding career in data analytics and generative AI. The link is given in the description box below and in the pin comments. Before we dive in, here's a short quiz question to test your understanding. If a company takes the blame for data leak, which ethical principle are they following? Profit, speed, accountability or marketing? Let me know your answers in the comment section below. >> Today uh we going to talk about uh data ethics. So it is all about foundations of ethical data practices. Like when we talk about these data practices uh it is all about working with data for different uh purposes like you may use data for some data analytics. You may use data to compute some data science activities, you may use data to train your machine learning model. So you know like uh people can use this collected data for different different purposes even like it can be used for targeted engagement and different different exercises. So this is like where we talk about what should be the ethical data practices while working with uh data or database projects and all. So whenever you install a new application on your phone, whether it is an iOS phone or Android phone, you know, uh this application is going to ask uh to grant certain permissions like permissions related to access your location, permission related to access your camera, your microphone or maybe sometimes your contact details as well. And these permissions are being asked by the applications to provide certain functions which are required for that applications. Uh the question is why do you think these uh phone or these iOS applications need to obtain your approval before accessing certain features? What might happen if these permissions were not necessary guys? What what do you think like what can be the consequences if we do not uh uh put these kind of restrictions or if we do not uh make uh user aware about the access of these uh uh different features of your phone services of your phone guys. So let's suppose there is an application and it is not asking your permission but it is still using your contact folder where there is a long list of your u personal data can be exposed. Yeah that that that's that's one of the consequences that we can think of you know your personal data will definitely get exposed if you uh you know if you do not make it transparent if you do not make it uh user aware access of that. Okay. What if let's suppose if if if application is tracking your location consent not taken from user. Yeah, that's what I'm saying. Let's suppose the application is not taking any consent but still it is accessing these part of information. So you know uh when you try to access any kind of these information for the user point of view this all comes under ethical data practices. Let's suppose if application is accessing your location um information then it can be used for surveillance uh practices by by the applications and if the user is not aware about that the location is being accessed by the application this is wrong like you know you can't be tracked by anyone else just randomly until and unless you are not giving your concern you know nobody can track you nobody can put surveillance over But as uh you know nowadays we are mostly um what you can say exposed with all these kind of digital applications. So it has become very important to understand uh what could be the consequences of sharing your personal information from the customer point of view as the same point if let's suppose I am an application builder now what should my accountability what should my responsibilities whenever I am accessing these personal information for my customers so this is like something we going to talk today and this is like where we try to understand like what is data ethics ethics and what should be the uh the right way to implement data ethics practices in the in the digital world and kind of thing guys. So uh let's quickly um go through with the learning objectives of today's sessions means at the end of these sessions I expect that you guys will be able to implement certain standard theoretical practices and legal frameworks and organizations to build customer trust and comply with regulations. That means you first need to understand like uh what are the ethical data practices and what are the different legal frameworks that you can impose or you can include in your organizations. So through which you can actually build trust between customers and you can uh comply the different regulations governed by the government. Then uh at the end of this session we expect that you should outline the importance of obtaining informed concern from individual group for collecting processing or sharing any kind of personal data. This is very important because in digital world you know you can't share any personal information any customer or individual personal information to anybody else without any concern. So it is very important that if you are collecting or processing any personal information you need to take the consent of the uh required person into that process and finally at the end of the sessions I expect that you guys will be able to create solutions to address ethical issues also that means uh let's suppose uh if somehow uh there is a there is an error in your system or some data breach has happened how you can address these issues. If the customer is is raising any issues, it is filing any complaint like how you can address these issues where you can you know keep yourself secure by what kind of data practices that you should include in your organization so you can keep yourself more secure into that process. So let's start with introduction to data ethics and this is the first thing that where we need to understand like what is data ethics and what comes under data ethics. So uh we can say that data ethics basically address different questions. Uh the questions regarding uh the different ethical principle guidelines and the standard governing for the collection use storage and sharing of data. So whenever you are collecting the data what are the principle of collecting a data using of a data storing of a data and even sharing of a data and I would add one more thing that is a processing of a data like how you can process that data also is very important. So data ethics basically covers all kind of um practices all kind of standards rules and regulations that govern these practices with the data like how you should collect the data how you should use the data how you should the data and if you are sharing the data with any third party what are the rules and standard to share the data. So data ethics covers all these kind of uh you know uh what you can say the rules and regulation and standards to be there. Uh now when it comes to the data it requires decision maker to consider impact for their decision on individual society and organization. So you know that is where like we collect data and this is something we have understood from our previous sessions also that we do uh data analytics uh because we want to convert your raw data into information and once you get this information you know you can use this information for making better decisions. So this information is basically help you to take better decisions. You can simply say that these decisions are better decisions. So that's what the ultimate goal of u you know of any organization who is doing data analytics or who is dealing some data science project or collecting data for that. Now when we do this as I said like uh these these organization need to follow some data ethics rules data uh ethics standards. So it basically talks about insurance or ensuring the privacy of data, transparency within a data, fairness within a data, accountability in the data handling process. When I say privacy that means if you are collecting a data that the data should be securely stored onto your organization and no one any unauthorized unauthorized user should not access on that particular data. When I say transparency, it's talking about uh uh the data should be used only for the intended purpose. So let's say if you if I am I'm collecting a data for any individuals for saying some X work that should be used only for the X work. The data should not be used for other purposes. That is where we talk about transparency. And here we talk about the concern of the customer or the end users to collect the data uh while uh while processing them. The fairness is that the data should be used without any bias into that and that is where like you know uh a fairness and we will talk more detail on that. Finally talking about the accountability which uh which is the uh role of the organization. So if there is any data breach has happened or any inconsistency of care in the whole process then your organization should make accountable for that guys. So these are some of the standard um you know practices that you need to uh study and you need to understand under data. When we talk about some legal and uh regulatory frameworks and uh so around the world since this digital data has become very popular and it is a become a way of you know uh sharing or storing the data. So there are some regulatory framework are being developed or you know created by different uh countries even some organizations are also involved there like a federal coselling and the US federal government. So organizations need to obtain certifications or they need to follow the rules and regulations provided by them while this data practices and data processes guys. So let's first understand what are the key principles of data ethics. um you know as I said like the very first principle is about transparency which means that you have to ensure openness about data collection method storage and sharing practices. So uh whenever you go to collect the data you need to be very transparent as I said like if I'm collecting the X information of any user you know I have to uh make it very transparent that how I am going to collect this information where I'm going to store that information what will be the uses of that information and if that information is being shared by any third party then what are the sharing practices then you need to be very clear while collecting the data you need to be very transparent. Second is a fairness. That means it avoids any kind of bias and discrimination in data collection analysis proc algorithm. That means if you are collecting a data, let's suppose you are collecting a data for um you know customer segmentation. You want to conduct an study for a customer segmentation and now you are collecting a data for that. So while collecting a data you know you need to touch upon every section of the customers. It's not that you are only targeting for male customers or you are only targeting for young customers or you are only targeting a customer from a specific demographic journey. If you're doing these things you means you are introducing some biases into the data and when you are introducing these biases that means you are discriminating other uh factors of you know the data to be take participation into that. Another examples that we can think of let's suppose uh I am building a model for rumé scrutinization. Okay. So let's say uh there is a model which basically uh process the rumé and then it simply give you that whether this particular candidate will be a good fit for interview or not. It's simply a resume a rum scrutinization application that we can think of that. So uh it it will give you whether it is a fit or it is a no fit. Now to develop this kind of model if I I I have to collect the data it should be like that I am collecting a data from you know equal. So it's not that I'm collecting data from mostly a male candidate. If I do this that means I'm introducing bias towards male category compared to a female category. And automatically if you go with a biased data to train a model your model become biased towards male category. And this way you know that that introduce discrimination. So whenever there is a female rum will come for a scrutinization it will mostly go for the no fits on. And this is where like we need to be very uh very much careful while collecting a data that the data should be collected with full fairness and all. Then talk about accountability and that's very important from organization point of view. So this holds individual and organization responsible for their data practices. Someone should be there that can be addressed if there are any uh breach happened or if there any inconsistency happens with the data analysis process or data processing practices and all that. Privacy it is basically used to protect individuals right to control over their personal information. So and every organization and everyone should respect this. So if the customer is not ready to reveal the personal information you cannot enforce him, you cannot force him to share that. And finally the security. So once you collect the data and you store the data now it has become your duty that you put some safeguard or some security measures through which an unauthorized access cannot be happened on data or that data cannot be manipulated by unauthorized user. So these are some key principle of data ethics where we talk about uh different dimensions we need to uh you know look for and we need to uh consider all these dimensions with equal importance. Uh let us understand this with a very uh very common scenario. You must have all used Snapchat guys. It's a very popular application. So what is this? This this this is an application which basically take take your face and then it try to uh play with the face like by putting some props on it or maybe try to change your face uh to a different uh what you can say different facial expressions and all this. So it try to add some creative filters that can enhance your selfie with facial recognition technology. So this is the main application you can think of that Snapchat uh basically for pop. Now while uh doing this it has to access your face. So you know it has to access your camera first. So when you put the camera to your face then it will take the facial image for that. Now it is going to use your facial data. Uh that's very important that is going to use your facial data for applying all these you know features or functions on it whether uh that can enhance your selfie or you can add some props on your face and anything whatever you do that but at the end of the day you need to give permissions to access the camera and then it'll going to take the image of your face. So now when uh we talk about such applications what could be the ethical considerations while doing this. So very first thing is when we talk about transparency as we have just seen what are the key principles of data ethics. So very first is the transparency that means Snapchat has to clearly explain the facial recognition technology purpose and give user control over their privacy setting. That means whenever we are releasing such kind of features in our applications, you know, you have to clearly explain that what is the technology that they are using to play with your facial image and over this they have to give the control power that whether a particular user will give consent to use that face for these purposes or not. So that is like it give full control over the privacy settings. So if user is not interested he can deny the access of camera for this particular application. This is like where we talk about control permissions and you know like whenever you install uh Snapchat kind of application on your system on your phone and when you use it first time it ask for your permission that you know this application is going to access your camera are you okay with that or you have to give the concerns for that. The second is fairness you know. So when you talk about the fairness it test algorithm for accuracy across diverse demography to prevent bias and discrimination. So it's not that if you are adding some enhance you know some selfie or some creative filters on on the faces. So it's not like that it is only working for some specific faces maybe some you know some fair faces and it fails on some black faces kind of. So it's not that uh like that if the application and algorithm is doing such kind of behavior that means the application is not fair with respect to the the audience with respect to the different customers of different breed different demography different races and all this. So that should not be the there and that has become the responsibility of the Snapchat makers that the application should equally perform on different race faces or different kind of demographic individual faces and all this. Then accountability. So it has to establish a dedicated team to promptly monitor ethical implication and address user concern. This is something I said earlier you know. So you you need a point of contact. There should be a point of contact which can address any data ethical issues when these are being raised by the customer or individual talking about privacy. So it's it's a very standard that user understand and control how their facial data is collected and used. So whenever this application is collecting any user facial data uh user should be very well aware about that how the data is being collected and how it is being used. So whether this should be only be used for some facial enhancement, some selfie enhancement or that particular data can be used for other purposes of that. So maybe u you know down the line because Snapchat is going to have u the data of millions of faces can snap can be used to train other models for that. So that should be you know very much clearly defined to the end user. And finally talking about the security. So uh the snapshot has to protect facial data with robust security measures to prevent unauthorized access. It should not be that the data can be stolen by third party or then it can be misused by someone else. So these are some uh standard practices what we listed on the previous slide like we have to maintain transparency while collecting the data. We have to use the data with full fairness. If you are using any algorithm behind this uh behind the application, that algorithm use the data with full fairness. There should be a point of contact who is accountable for any any kind of issue raised by customers or any data breach happened. Uh the data should be um collected in a complete private manner with user consent and the data when collected should be safeguarded with proper security measures. So any unauthorized user cannot access the data. I hope we are clear with all these points guys. Now we do we understand like what are the key principles of uh data ethics? Transparency, fairness, accountability, privacy and security. So uh by understanding this let us try to understand why it is important like why we talk about these key principles why we talk so much about data ethics rules and regulations in uh in data practices and all. The very first important thing is that it help you to build trust with individuals you know. So whenever a user is connecting with you, whenever user is sharing his data with you on any platform, if you come up with all these data ethics principles, it will help you to build trust between you and the customer the entity. Then it protect individuals and society from any kind of data breach. So if any organization is implementing all kind of data ethic practices, it basically help individuals as well as society for any kind of database. That means if you have collected the data and you are storing this data onto any database, you kind of you know adding some security safeguard. So any unauthorized user cannot get access on this data. And this is where like user feels secure that my data is being protected and there should not be any kind of database happened over there. So when you implement these data practices a user get confidence on that and even the data is being protected by any kind of database. Third uh it promotes responsible AI developments. That means uh whatever the AI model are being trained on the data if you are collecting that data with full fairness and transparency that means it promotes the responsible AI development. The model will not come with any kind of bias. The model will not do any kind of discrimination for a particular type of individuals. So that also help you to do uh responsible AI development and finally it contribute to a fair and equitable digital future. So everybody is getting equal chance for uh you know move ahead. So that help you to contribute a fair environment where uh we can move into a digital future. So with this importance let's talk about significance of data ethics and data index. So you know like I said uh the the primary objective of data analytics is to convert your raw data into but in this process also you know we have to collect the data we have to process the data we have to store the data somewhere and then sometimes you have to share the data also. So again data ethics become very much important in the process of data analyics also. So talking about significance of u uh data ethics and data analytics u you know data analytics is what it basically unlocking insights and innovation from your data that's the primary objective and um when we define it it reveals the hidden patterns and trends within a data that empower anyone to take decisions like it can help you to make informed decisions. This is I think we are all well understood that what is the objective of data analyics and how it help you to take well or better informed decisions like in that process. Uh second it optimize process and performance. So data analysis identifies the improvement area. That means uh when you do these analysis uh you can definitely know what are the gaps that you need to fill. Uh what are the shortcomings in the process that you need to address. So in the analysis process now you you you will be going to get all these kind of information. So that help you to definitely optimize your existing system in terms of process and performance. And finally u the outcome of the data analytics once you identify any patterns once you identify any trends or associationship in the data you know you can certainly use this information to train a model that can help you to make future predictions. So these are the basic uh you know uh the objectives of overall process of data analysis. What we say you know data analysis help you to make better decisions uh make more informed decisions. Uh data analytics can help you to optimize your process and performance of the system and data analytics can help you to make future uh predictions in terms of some trends and behavior that you can see from the past data. Now with this process when you introduce the ethical considerations you know in this process we talk about bias and discrimination and this is something I said it earlier also. So when you uh train any algorithm when you train any model on a biased data you know u that model will produce biased result and when it produce biased result that introduce discrimination into that. So the if you do not follow uh data ethical principles in data analytics process you may end up of getting a biased model. A simple term we can say it's a biased model and biased model is going to introduce discrimination in the process. Uh privacy concern is another thing. So data collection and analysis can infringe upon individual right of privacy if not handled responsibility because uh you know use cases such as like customer segmentation or use cases such like like you know Facebook friend recommendations you know you have to collect certain personal information of individuals the demographic detail sexual detail the age um even political opinion you know the interest rate the hobby and all these informations are considered to be personal information if you do not handle them properly these informations can be used in other bias then that you know you can simply impersonate a person. So a digital entity can be created which is going to going to impersonate a real identity. So this is very important to have uh data ethics in the process of data analytics. lack of transparency. If you are not very clear you know with your data practices that can you know uh introduce some uh errors and that can you know dent your trust and hinder your accountability also. So data ethics uh practices are very important in the process of data analytics. uh this is one of the uh important aspect as I said like you know bias and discrimination is very important because if your data is not collected fairly uh it will come with a bias and when you train any model onto a biased data it is it the the predictions will also going to be a biased guys and that will introduce discrimination. So the very first thing is that when we collect the data we avoid biased data and algorithms. Addressing bias in the data and algorithm is crucial to prevent discriminatory outcome and promote responsible use of them. That's very important and that's very important concern whenever we build such models that are specifically built for a community uh that are specifically built for uh societal applications kind of thing. Uh while doing this we ensure algorithmic accountability also that means we have to create mechanism to hold algorithm accountable ensure fair and responsible datadriven decision-m and this can only be possible when your data was collected fairly and each and every section is uh selected randomly or getting an equal uh equal weightage while by while by while by while by while by while by while by while bywhile by while by while by while by while by while by while by while bywhile by while by while by while bywhile by modeling process uh that also promote diversity and inclusion of data teams so identifying and addressing biases within diverse data analysis team ensures fair enough. So if you find that you you see one category is more than the other category then you can form teams that can help you to introduce more data points and try to balance the data. Uh this is like where we can handle the bias and discrimination guys. Uh so let us talk about one of the use case and that use case is related to one of the feature of Facebook uh which is a targeting advertisement. you can call it as targeting advertisement. Now here if we uh consider uh one use case of discriminatory practices uh that means uh in the Facebook is is um you know showing a particular ad to a particular community. This is what we call a discriminatory practices. So what could happen? So there uh there is an incident let's suppose we listed for 2019 that Facebook faced criticism over a targeted pract targeting practices allowing potential discrimination based on gender and age toward a job advertisement and this is actually uh this can be possible like so you can filter and you can say that if I'm uh I'm issuing or I'm uh you know uh displaying an advertisement uh that advertisement should be visible for a particular gender and a particular age group. Now this is where like uh we talk about discriminating practices in that sense like other people will not get the recommendations for those job openings and this is where like they feel um you know discrimination so they will not get chance to get appear into that. So advertisers could restrict ads to a certain age and gender demographic. Now when Facebook will go with such kind of practices that means it is introducing discriminatory practices and this has happened with the Facebook in the 2019 it has you know received lot of crit criticism over that but this is one of the feature of that application guys. So when we use Facebook for social advertisement, we can you know add different different filters. You know you may add those filters based on gender, you can add those filter based on age, you can add those filters based on certain demography. You know what these these are the different features that uh Facebook uh has and you can apply these filters. So uh the advertisement become targeted advertisement. You can say it's a targeted audience advertisement. it it will only be visible for those many users which are passing these filter criteria and all this. So this this become a discriminatory practices because it has been raised in 2019 by somebody that you know this is not right. You are not uh fair while providing these uh these advertisement or you know issuing the information to your end user. So the ethical concern in this whole practice is that targeting a specific demographic could lead an unfair exclusion from the job of opportunities. That's very common thing. You know when you do this the other people who are not coming from the same demographic reason may not get an equal opportunity to uh get selection or even to apply for the job. Now exploiting ad targeting tool could undermine fairness in the job market. Lack of transparency, raised accountability and trust concerns. When it goes um you know your user, your end user who is the customer you know will not be happy because sometimes you feel that you know you have been discriminated by the platform that particular ad was not shown to you and in such cases you may lose trust with the customer and there will be some trust concern and you may lose the customer as well over here guys. So that is where you know as ethical concern and from this what lesson is being learned. So uh you have to put some regularly audit algorithm to mitigate the biases. So you know you have to identify like how you can make it better how you can uh prioritize the ethical design for ad targeting rules. So whenever we talk about such kind of filters these filters should not be very harsh or they should not be in the practice. If you are flowing an ad that should be flown to everyone and everyone should get equal opportunity to see that ad guys. So this is the very first thing where we talk about uh one of the scenario of bias and discrimination. The second thing is like we talk about privacy uh no when we talk about the privacy concern. So privacy concern is mostly related to uh the personal information of any user. So the best practice that you can use is data minimization and purpose limitation. Okay. So always like whenever you are implementing data ethics practices to any of the applications uh you should always try to uh uh collect minimum information for an individual. So this this is what we call a minimizing data collection. Okay. And you should also store the minimum information. So the information which is not being used for a longer period of time you should not store them. You should collect them instead use it and throw it. So that is basically help you to uh reduce any kind of security risk or any kind of privacy risk to be involved. So that's a very good thing uh that is being introduced means you only need to collect the piece of information which is required for your application. you should not u you know collect more information about the user and that's what we call data minimization and purpose limitation. The second is uh to to ensure the security in your system you have to install some robust cyber security measures. So only the authorized user will get the access of the data and no unauthorized user can access the control. So that's another practice but it is outside of the data collection process. is something related to the organization point of view and there are some clear data breach protocol are also we introduced that means if there is any data breach happen how you going to address them you have to be very clear with the process so that's what we call data breach protocol and when you display these data breach protocols it basically help you to build trust and safeguards with individual during security incidents um one of the use case is with T-Mo databach uh that is something we have seen in the past. You know T-Mo is is a major mobile carrier company. You must have heard this name guys. And that has been suffered with a big data breach in the January 2024 where some hacker gained unauthorized access to the application programming or API access and it stole the information of around 37 million customers and this is a big data guys. So when you become a customer of any mobile service company you know you put lot of detail you put lot of personal detail of you. You put your name you put your surname you know you put your address and over that uh there is a phone number the mobile number associated with your name and these are all uh personal identifiable informations we call them PII. So in that sense you know if any person has got the information of 37 million customers he's going to be a king in the data side guys we can use this information for any purpose you know so this breach like what what they learned from this particular incident happened in 2024 this breach expose sensitive customer data putting them at a risk of identity theft fishing scam targeted marketing you know uh the the hacker can use this data for any other purposes like it can apply fishing scam. This is very common in India. You know if you see like lot of people call you you know they they try to give you lot of offers or sometime threat as well and then you know you they ask you to share your accounts or CBB numbers etc etc kind of thing. So uh they can be used in a fishing scam. Uh this can be used for the targeted marketing also you know. So this is like where when when you expose such important information to an hacker community that can be used in any purpose. So the lack of proper API security measures raise ethical concern about the company's handling of sensitive data when it happens. You know obviously the customer uh trust will be lost in the whole process and customer may not rely on T-Mo for them. So what is the lesson learned in this particular incident? So firm must enforce a strong API security. Obviously the the the weak interface allow unauthorized access and in that sense you may lose the customer information. So you have to enforce a strong API security measure that include access control breach detection and ongoing monitoring and whenever you know any kind of breach happened it should shut down the system as well. If the breach late discovery highlight the urgency of promptly detecting anomalies it virtual companies clearly communicate with the customer about breaches and remedial actions. So that need to be taken care of. So that's very uh theoretical I guess you know. So whenever we talk about these incidents we'll take uh lessons from them and try to address the the issues over that. So here like we talk about API security I must say that's important. Okay. The second is uh we talk about timely detection. So we should have a method and third is we have to make transparency also to to make your customer aware about these kind of breaches and how you going to address them remediation actions and all this. Any question here any any other scenarios which is coming to your mind? Do you do anybody know about any other kind of incident uh where like we can see the the these kind of use cases where we have seen the breach of data or what is the ethical concern related to that data and what are the lesson learned from that particular incident. Does anybody uh know any incident? You're all working in the industry. You must have seen that. Okay. Okay. So now let's move to the third which is a lack of transparency guys. So you know like very much is the bias and discrimination then the privacy and third is a lack of transparency. Uh transparency is all about providing a clearcut information about data collection and the data analysis and the uses of data. So basically we can understand and identify the different biases. So you have to uh tell the user that what type of information you are collecting, how you going to analyze that information, what is the use case of that uh information and where you going to store that application uh information. So this is in that way you can make your uh system transparent. But when we say lack of transparency that can introduce some you know risk or um uh some issues in this process. So let's suppose if we talk about algorithm transparency it providing a clear explanation of data collection analysis and use understanding and identifying the biases. If we talk about risk and limitation of the communication the transparently recognized data analytics limitation and risk like biases foster trust and informed decision making and user control over the data that means you are providing uh control over their data and ensuring access correction and delegation rights uphold privacy and prevent the users. So uh when you introduce the transparency at least you have to address these things. Your algorithm should be very transparent in terms of how the data was collected whether there is any bias in the data or not. And uh what are the limitations or the risk involved with the communication that should also be very clear and finally you should give the control to the user that if he's not interested he can actually uh you know uh erase the data he can modify the data he can alter the data. It is not like that once the data has been collected the user has no rights or no control over the data and uh one of the uh one of the use case related to the lack of transparency is with uh Google incognito mode guys. So I hope you have seen this incognito mode of Google Chrome. What is the purpose of that? Why we use it? Why we use incognito mode? How it is different from the normal Chrome? No cookies saved by bidding some info. Yeah. No history, no password save. So that means whenever you do browsing, you know, whenever you browse anything, you know, it it remain anominous, you know. So we believe that whenever we do any kind of surfing, any kind of searching, any kind of browsing, it remain anominous. So uh nobody is going to know that who is act who is browsing that information kind of because it is not tracking there are no cookies stored into your local system no password saved no streaming recorded and in some way you try to maintain your privacy but what if you know what if if uh you got to know that Google is recording that piece of information while you are working in on incognito mode. Okay. So that's something you know that's something is one of the scenario. So millions of Google Chrome user relied on incognito mode for private browsing. That is what we use for. So our private information should not be shared over the browser. Uh the location uh the passwords the personal history nothing nothing is going to be shared. That is something you know we rely on that in bug mode. However, you know there has been a lawsuit alleged Google to continue track user activity even in this mode and creating a privacy controversy. So again there was a challenge given to Google that even in the incognito mode you are tracking the user activities like what kind of browsing it is doing from where it is doing even though we were relying that if I am in incognito mode my private information is not being shared to anyone else. Yes. So that is very important and in that process if you go with lack of transparency you know this is what a lack of transparency. The very first ethical concern is the deception you know that is Google mislead user about the true capabilities of the incognito mode. So Google has not made its user very much aware about that even in incognito mode there are some pieces of information are collected. There are some users activities tracked that is what we call deception. Now data privacy the users were expecting uh that their information is completely private in the incognito mode but it is violated because the the Google was tracking the user activities and in that process you can say the Google transparent uh transparency was not good enough for data collection practices that means it was lacking in the transparency guys. So what are the lessons learned from this that user trust rely on transparent communication and meeting expectations. So if you are claiming any services if you are fulfilling that services then only you can say that you are able to you know be uh being able to earn the trust and you became a full transparent understanding the limitation of incognitive mode and exploring better privacy options can help make informed online decision. So that is you should make aware and then in that way like you can you know you can make your u users more informed and accordingly it can use the uh the incognitive mode and then if Google is collecting that uh data that is okay but that can be used for non-purpose or transparent way guys. So these are the three major areas where we talk about the practices of data ethics in data analytics. One is related to the bias and discrimination uh discrimination. Second one is uh the privacy and third one is the transparency guys. So now uh we can talk about ethical analytics. Ethical analytics means data ethical practices in data analytics for better future. The very first thing is we want to build trust technology or trust in the technology. So when you enforce all these key principles you know accountability, transparency, privacy, security and fairness, these are the five principles of data ethics. When you introduce them into a data analyics process, it help you to building a trust in the technology. Ethical data practices build trust. Fostering responsible technology advancement promoting inclusivity and social justice. So when you remove biases in the data process that introduce fairness and that is good for the society. Shaping responsible AI development. I think this is something I have already explained you. When you introduce a fair data in the model training your model will give you fair judgment. If you introduce biasness in the data your model become biased to a particular category which is not good for the society. So when we talk about uh introducing data ethics in data analytics process especially in developing models it will help you to shape responsible AI development best and then safeguarding individual rights and freedom. So when you introduce uh privacy rules when you introduce security enough security that help you to safeguard uh individual personal information as well it also preserve the individual rights of the information right so if the user is ready to share his personal information that is okay if it is not share that is also okay for it uh let's have a quick recap on uh whatever we have discussed guys it is more of a theorical so which principle emphasize collect collecting only the necessary information for a specific purpose reducing sensitive data and storage. One is transparency principle, data minimization principle, cyber security principle or explanability principle. What do you think what is the right option? Yeah. So you can say uh B is the right option that basically u talks about collecting only the necessary information. You should not collect the information which is not required for the application or for the use case that you are solving. In that way you only collect minimum information and you store minimum information. So you know that's not good to store sensitive data in a use amount because if any data breach happens and that way you lose the trust of the customer guys. So right option is B. That's correct. Now let's talk about uh ethical challenges in data analytics like uh so we have seen like what are the common key principles. Uh so the common key principles are like you know you talk about transparency you talk about fairness you talk about accountability you talk about privacy and when you introduce these data ethics into data analytics you know you talk about three major things. One is um wiseness and discrimination. Second you talk about uh the privacy concern and third you talk about lack of transparency. This is something we have seen till now. Now when you introduce them like you know so what are the challenges u what are the ethical challenges that uh you see like when you introduce these things into that let us discuss them and try to understand them in detail. So uh let us suppose uh we take a scenario uh where there is a mobile application that tracks user location and that location data provide personalized recommendations and advertisement. So remember what application is doing? It is actually collecting the location of the user means demographic information and this application is using this information to provide personalized recommendation and advertisement. User are unaware of the extent of data collection and have limited control over how their data is used. So that is the current scenario. So application is not giving the access or the control to the user. So it can actually uh modify the data and there is a only limited content uh limited u control is given and even uh the user is not very much aware that how much extent the data will be used or how it is being collected guys. So with this scenario what do you think what are the ethical concern associated with above scenario or if you look for that what are the lessons that you will learn? Very good. One you can say the privacy if you do not have a uh full control how the data is used you can say the privacy is one of the concern personal data is being used that's right can you say there is a lack of transparency also available here if you do not know that how much extent the data is being collected and it is being used when I say that transparency is also not very much available here. If you are using this for a personalized recommendation advertisement what is the issue related with that security also involved you are putting a surveillance kind of system you are tracking the personal location. Okay. So with that what do you think if these are the concern what are the lessons learned from this scenario what we need to put what we need to introduce in our organization what we need to introduce in our application process very first we need to put a very clearcut transparency policy that tells about that we are collecting this amount of data from you and this data will be used for that purpose only then you need to have a proper privacy policy also. Okay. So there will be a clearcut transparency and privacy policy. Then you have to provide a proper security mechanism to safeguard the data and you have to provide a proper control over the data as well. So if the user is interested it can actually change its information and settings. So going with scenario we can say that it is actually violation of privacy rights lack of informed consent potential for targeted surveillance and manipulation. So these are some ethical concern that are listed here and what are the lesson learned? So need to for a cle