Hi, I'm Mrs. Allchin, and I'm going to be taking you through the citizenship lesson today.

I'm going to give you all the information that you need to be able to take part in the lesson, and I'll also pause and tell you when you need to complete an activity or complete a check for understanding.

I hope you enjoy the lesson.

This lesson is entitled "A Case Study: India, what is an AI election?" And it comes from the unit, "How is social media changing our view of democracy?" By the end of this lesson, you'll be able to critically evaluate the impact of using AI in a general election.

These are our key words for the lesson.

So artificial intelligence, AI, which is technology that makes machines think and act like humans.

Manipulated, which is controlled or influenced, often unfairly or dishonestly.

Analysis, which is the detailed examination of something, breaking it down into parts to understand it better, and tactic, which is a planned action to achieve a specific goal.

So this is our lesson outline.

We're going to look at how was AI used in the Indian election.

We're then going to look at what were the advantages and disadvantages of this, and we're then going to consider should AI be used in future elections, And we're going to start off by looking at how AI was used in the Indian election.

So John is asking, "What does AI stand for again?" So if you want, pause and think.

What is AI? What do you know about it? So AI stands for artificial intelligence, and it can be used in digital media to learn and make decisions almost like a human.

It can do things like recognise faces and photographs, it can talk to you like a chatbot and even be used for self-drive cars.

So it is a really, really fascinating invention that has the capacity to do lots of things.

AI can also be used to create deepfakes, which involve manipulating a video clip or image to create new deceptive content, often with known faces or voices.

And according to the BBC and other news sources, in the 2024 Indian general election, AI was used frequently, and it played a significant role in both political campaigning before the election and voter engagement during the election.

And the AI methods that were reported as being used included deepfakes, social media analysis, robocalls, chatbots, disinformation using AI content, and translation features.

And we're going to look at these during this lesson.

So John is asking, how were deepfakes used during the Indian election? Let's have a look.

So political parties use videos which were manipulated by AI to look as though deceased politicians, so politicians that had died, were supporting current political candidates.

This was a highly emotive tactic, and what that means is it's a tactic that really got people feeling passionately, and strongly, and really kind of resonated with their emotions.

And it was highly emotive as the familiar and often much loved faces of past leaders that were no longer with the people in India delivering messages of support could build emotional connections with the voters themselves as even though the electorate knew these people were deceased, hearing their voices, even through AI, could be hugely impactful.

So let's have a quick check.

Deepfakes of past leaders endorsing candidates were impactful because they were A, emotive and familiar, B, energetic and funny, or C empathetic and fictitious.

Have a go.

And it's a because they were emotive and familiar.

So Jun is asking, "How could social media analysis be useful during an election?" So have a think yourself.

How could using AI to analyse what people are posting and talking about on social media be useful during an election? So AI can be used to analyse voter opinions by using what is being posted and shared on social media accounts.

This gives political parties information about who is popular, which policy citizens are agreeing or disagreeing with, and even voter intentions.

Political parties can then use this analysis to tailor their campaigns and target specific information to the right people.

And this tactic was used heavily during the Indian election.

So let's have a check.

True or false.

Social media analysis has no benefit or use during a political campaign.

Is that true? Is that false? And can you tell me why? And that's false.

Why? Political parties can use social media analysis to tailor their campaigns and target specific information to the right people.

So he's now asking, "What's a robocall and how were they used in the Indian election?" You might never have heard this term before, but just look at how the word is kind of put together, and just see if you can have a bit of a guess about what this might be.

So a robocall is a phone call that uses a computerised auto dialer, and that software that automatically dials phone from a list to deliver a prerecorded message.

During the Indian election, robocalls were frequently used, which allowed politicians to interact with many citizens also in multiple dialects, so different dialects in different languages, and in real time.

This aimed to massively increase outreach as a robocall were automatic and didn't rely on a person manually calling thousands of people.

Because if you can imagine if you worked in the office for a political candidate and you had to manually phone and speak to hundreds and thousands of people, that would take absolutely forever.

Whereas if it's done with AI, it's really, really, really quick, and the AI can also make sure the correct dialect and language is being used as well.

So what about a chatbot? How were they used in the Indian election? Again, pause and have a think.

See if you know.

So a chatbot is an AI computer programme designed to stimulate and mimic human conversation with internet users.

AI is used to make the back and forth conversations read so they have come from a real live person, rather than an AI programme, which could increase emotional connections and engagement.

Chatbots we use frequently through the Indian general election, providing the electorate with information about registration, polling locations, and policy queries.

So you might have even come across something like this yourself, or your parents or carers might have come across this.

So this is when you can type and find information, and it sounds and looks as though you're communicating with a person, which makes it a bit more personal when actually it's a chat bot and it's all very cleverly done through AI.

So let's have a check.

Can you match the AI method to its description? Which one is robocall and which one is chatbot? So they were actually in the correct order.

Hopefully you got those correct.

So we're now asking what's disinformation again, and how was it used in the Indian election? So the word disinformation may have been a word you've come across before.

Can you remember what does it mean and how might it have been used you think in the Indian election? So disinformation is false information that is knowingly spread to deceive and mislead others.

And during the Indian election, deepfake videos were used to spread disinformation.

For example, one such video falsely depicted the union home minister, and that's a person that's responsible for law and order, announcing changes to policies that guarantee a percentage of seats in government jobs, education, and legislature for citizens from marginalised casts and tribes.

This disinformation was highly emotive and could have misled Indian electorate.

So that could have got people that were potentially in marginalised groups within India panicking and thinking that things were going to happen to policy that was going to really negatively influence them.

So this is an example of how disinformation could be really misleading in an election.

Now Jun is asking, "How were translation features used during the Indian election?" So just think, what does it mean to translate something and how might that be used during an election? So first of all, to understand the impact of this, I think it's a good idea to think about actually India and the languages that are spoken there, because according to the Indian constitution, there are 22 recognised languages, but there are also hundreds of additional dialects which are variations of traditional languages.

So the citizens of India between them speak lots and lots and lots of different languages and dialects.

So during the elections, AI created translation features which were used so that the political messages could be understood by all citizens, regardless of the language or dialect spoken, and this increased the reach of campaign messaging.

So let's have a quick check.

Which is an example of disinformation that was used during the Indian general election.

Is it A, B, or C? And it's C.

Deepfake videos of deceased pass leader endorsing candidates.

So for task A, I would like you to write a brief description of each of the AI methods used during the Indian general election.

I would like you to include deepfakes, social media analysis, robocalls, chatbots, disinformation using AI content and translation features.

And you're being reminded that you don't need to evaluate these.

You just need to literally say what they are and describe their function.

So your brief description of each of the AI methods used during the Indian general election may have looked something like this.

So deepfakes are manipulated video clips or images to create new deceptive content.

Social media analysis analyses political opinion on social media.

Robocall are phone calls that use a computerised auto dialer to deliver a prerecorded message.

Chatbots are a computer programme designed to stimulate and mimic human conversation with internet users.

Disinformation using AI content is when knowingly false information is spread via AI methods, e.

g.

deepfakes, to mislead.

And translation features translate information into multiple languages and dialects.

We're now going to move on to what were the advantages and disadvantages.

So we're going to look at these methods in a bit more detail and think about what was positive and potentially negative.

So Jun is saying, "On the whole, I think the use of AI in India's general election was positive And Alex is saying, "It's not quite as simple as that.

We need to think critically when exploring whether the use of AI during the general election in India was a positive thing or not.

Although there were clearly some advantages, there were some major disadvantages too." And that's really important when we're thinking about citizenship skills, that we're really thinking critically about both sides of the argument.

So let's think about the advantages.

So advantages of using AI in the Indian general election could include, it increased and enhanced voter engagement, it improved and increased outreach, so getting out to more people.

It allowed for multilingual information.

So that's what we're talking about, the translation services, and it improved efficiency, so it made things quicker.

So Jun is saying, "I agree, what do you think?" So what do you think? What are your initial thoughts about this? Was the use of AI a positive thing during the election? So during the general election in India, AI was used by political parties to create personalised messages that targeted the specific concerns that different communities may have had.

As in India is the seventh largest country in the world, AI allowed the right information to get to the right people, potentially supporting them with making positive and informed voting decisions.

So for example, unemployment in urban Delhi was an issue during the election as it often can be in major cities and countries.

As according to the Times of India, the rate of unemployment in urban areas of Delhi was around 6.

7%.

So political parties may have targeted voters in this part of India with information that directly addressed this concern.

And according to the Indian express, there was also concern that approximately 30% of employed people in urban Delhi worked within small businesses that often had limited opportunities for progression, so e.

g.

street vendors and hawkers.

So this is an example of there might be potentially people in certain areas that want to know specific information 'cause that's going to really drive their political behaviour.

So let's have a quick check.

I want you to have a read of this statement, and think what are the missing words.

So during the general election in India, AI was used by all political parties to create personalised messages, that targeted the specific concerns that different communities may have had.

General elections are vital in democracy as they allow citizens to vote for people to represent them in government.

The use of AI in India to translate political information in real time across the many different languages and dialects of the country was a huge advantage.

The most widely spoken language in India is Hindi, but it's not spoken and understood by all.

So during the election, AI was used to translate speeches by candidates in real time, so as they were actually happening.

And this ensured that all Indian citizens could listen and understand what was being said.

In such a linguistically diverse country, this was a powerful tool for democracy because regardless of the language or the dialect you speak, you absolutely need to be able to hear what potential candidates are saying and AI allowed for this to happen.

So true or false.

There was no need for translations during the Indian general election.

Is that true? Is that false? And can you tell me why? That's false.

Why? There are multiple recognised languages in India as well as hundreds of dialect spoken.

Although the main language is Hindi, it's not spoken by all, And this highlights the need for translated information.

AI also increases the speed in which functions can be performed.

So during an election, this can improve efficiency of tasks such as speech writing, data analysis, and outreach work.

Not only does this ensure that time is used efficiently, it can also save money on campaign costs too.

So let's have a check.

Which is not viewed as an advantage? Is it A, B, or C? So increased use of deepfakes is not viewed as an advantage.

Disadvantages of using AI in the Indian general election could include the following.

So it could allow for disinformation to be easily spread, it could bring up data privacy concerns and it could be viewed as manipulative.

So Alex is saying, "Actually, I agree.

What do you think?" So again, pause and just think on the whole what are your opinions of AI being a potential disadvantage in an election? So AI-produced deepfakes can make the spreading of disinformation much easier.

Deepfakes can look really, really or sound really, really realistic, and this can mislead voters and potentially impact voting behaviours.

In a democracy, citizens should be able to rely on the media to access factual information that helps them make informed decisions.

Therefore, the use of AI to create realistic fake content could undermine the democratic, and it can also lead to people not actually trusting the media.

So let's have a quick check.

What are the missing words in this paragraph? So pause while you have a read, and then think what they could be.

And you've got a little bit of a clue there because you've got the first letter for each of the missing words.

So in a democracy, citizens should be able to rely on the media to access factual information that helps them make informed decisions.

Therefore, the use of AI to create realistic fake content can undermine the democratic process.

Although social media analysis allowed for targeted campaigning, there were concerns by some that this could raise data privacy concerns as well as ethical concerns.

While some might argue that information published openly online on someone's social media platform is in the public domain, and therefore free to be analysed, others might argue that this is unethical as people's personal accounts are being used for political gain.

There are also concerns that AI could be taken further in the future and used to hack interpersonal information that isn't in the public domain.

Targeted campaigning using AI tools could also be viewed as manipulative.

This is because an individual's identity location or their their individual personal social circumstances could be used to try and persuade them to vote a certain way.

While some may argue that this is a positive step in increasing voter engagement, others view it as propaganda and feel it has no place in a democracy as everyone should be fed exactly the same information and be free to make up their minds accordingly.

So let's have a check.

True or false.

Targeted campaigning can only ever be viewed as an advantage for democracy.

Is that true? Is that false? And can you tell me why? It's false? Why An individual's identity, location, or social circumstances could be used to try and persuade them to vote a certain way, which some people might view as propaganda.

So for task B, I'd like you to think, what score would you give the use of AI during the general election in India? So if you were to score like a one or a two, that would be that you see it as having significant disadvantages and it shouldn't be used.

Whereas if you're putting it somewhere near 9, 10, that's significant advantages, and AI should be used, or you might be somewhere in the middle.

So justify your score using the examples explored at this lesson.

So your answer may have looked like one of these when you were justifying the score that you would give for AI during the general election in India.

So you gave it somewhere quite low, this one is a 2 out of 10, you might have said, "I think the use of AI during India's general election was hugely misleading for voters.

Using deepfake videos of deceased political leaders was emotionally manipulative, and accessing people's personal social media accounts is ethically wrong.

AI can lead people to distrust the media which is damaging in a democracy.

AI should only be used for translation purposes.

That aspect is useful." Or, you might have put somewhere in the middle.

And this might be something similar to, "I can see the advantages and disadvantages equally.

I think it is important that political information is targeted.

This allows people to find out about the policies that will impact them the most and lead to more informed decision-making.

However, I also believe that all citizens should read policies relating to all matters as this ensures greater understanding of a political party overall.

I also have concerns over AI going too far with data analysis and potentially hacking into people's personal information." And then lastly, you might have given a really high score, and it might have read something like this.

"AI is absolutely the way forward.

The election in India hugely benefited from this, in particular, by allowing citizens to equally access vital information regardless of the dialect that they speak.

This is a powerful tool for democracy.

I also think that political information should be targeted.

Citizens have the right to know more about the policies that will directly impact them.

If you were someone who was concerned about unemployment or job prospects, you would want to know what the different political parties were proposing to address this.

Whereas if this wasn't a concern to you, you would likely turn to other policies to help make your decision.

I think AI is great for getting more people engaged in politics, which is great.

We're now going to move on to looking at, should AI be used in future elections? So many people believe AI is a great invention because it can process and analyse huge amounts of data at incredible speeds, allowing for improvements in all areas that impact our day-to-day lives, such as healthcare, transport, and education.

And therefore a key argument is, if AI can benefit these areas, surely it should be used to improve elections too.

So how could AI have the potential to improve future elect elections? So it could be used to improve administrative tasks, which can take a long time when completed by human.

So for example, voter registration, counting the ballots.

And this could also remove the potential issue of human error, and also increasing engagement by targeting campaign material and also utilising AI features such as chatbots and translation features.

Whilst concerns regarding AI generated deepfakes are valid, this argument can also be viewed from an alternative perspective.

AI could be used to effectively identify disinformation, lots and deepfakes.

It could also be used to find and delete damaging online content in the run-up to an election, such as extremist content.

And this could reduce misleading and harmful narratives, 'cause we need to remember, yes, AI can be used to create deepfakes, but it can also actually be used to detect them and detect disinformation.

Let's have a check.

What could AI be used to identify during an election which could be of benefit? So you might have put one or even more of these.

It could be used to identify disinformation, bots, and deepfakes.

However, some people believe there is not a place for AI in future elections as the potential risk outweigh any benefits.

And some of these risks could include concerns around personal data and the risk of hacking, citizens only receiving half of the story due to overly targeted political information.

And deepfakes also being used to mislead the public.

There is also the issue of cost because advanced AI software can be really, really expensive.

And also for it to be really utilised to its best advantage, it needs data scientists and model engineers to ensure that it is used to its full capacity.

So this does not come completely free to be used at a really high level.

This is going to cost money, and therefore this raises an ethical question.

Is it right for larger political parties who can naturally generate more income to have better access to advanced AI tools and the advantages they bring? Because larger political parties are going to have more revenue and more money coming in than smaller political parties, and that does pose an ethical question.

So let's have a check.

Which is not classed as an issue in using AI? Is it A, B, C, or D? And it's D.

Administration for efficiency is not classed as an issue in using AI.

Some people may also question whether AI should be increasingly used for administrative tasks that are currently carried out by humans.

Although humans are slower, and there is a risk of human error, this needs to be balanced against the risk of large scale hacking and computer viruses.

Let's have a quick check.

True or false.

There is no risk attached to using AI for administration during an election.

Is that true? Is that false? And can you tell me why? It's false? Why AI runs a risk of large scale hacking and also issues with viruses, and it's a tricky one to balance.

Would AI be quicker? Yes, absolutely, but there are also risks involved.

So for task C, I want you think about, should AI be used in future elections? I want you to really think critically about this.

So explain your answers showing that you have considered at least one other point of view.

So you're going to state what you think, but also state what the opposing argument could be.

You may wish to reflect upon efficiency, manipulation, and engagement.

So this is what your answer may have looked like.

So on the one hand, I believe AI does have a place in future elections.

AI has proven to be of significant benefit in many areas.

So it is only fitting that it should be utilised to improve elections as well.

There is so much political information to digest during the election period, and AI could be used to ensure citizens have quick access to the information that matters most to them.

Though I can appreciate people's fears around deepfakes, I think it's important to remember that AI can also be used to detect and delete disinformation.

On the other hand, I can see how data privacy could be an issue.

With increased use of AI, clear steps would need to be taken to ensure our private data is safe from AI software.

I also think there would need to be an assurance that AI could be used to quickly identify deepfake videos to ensure that the public weren't being manipulated by disinformation.

So in summary of the lesson that was looking at the "Case Study of India, What is an AI election?" AI played a key feature in India's general election with deepfake, social media analysis, robocalls, chatbots, disinformation using AI content, and translation features being heavily used.

AI had some advantages, such as increased and enhanced voter engagement, improved and increased outreach, multilingual information being available, and improved efficiency.

But it also had disadvantages such as the spread of disinformation, data privacy concerns, and being viewed as manipulative.

Some people believe AI does have a place in future elections due to it being modern, quick, and effective in voter engagement and outreach, whereas other people believe it doesn't have a place due to the risks relating to privacy, disinformation, and cost.

Brings us to the end of this information looking at the case study of India and what is an AI election.

Well done for all your hard work and I hope that you will come back for some more citizenship lessons in the future.