Why will it be difficult to hold free and fair elections in 2024?

The rapid growth of artificial intelligence (commonly referred to as AI) is creating many opportunities around the world, but it is also raising serious ethical concerns. The United Nations is Interim Report: Governance of AI for Humanity The report calls for closer alignment of international norms and AI development practices through recommendations to strengthen accountability. The report highlights the importance of identifying, categorizing, and addressing risks arising from AI based on the principles of inclusiveness and public interest. AI technologies have the ability to generate and spread misinformation with alarming scale and speed. With elections coming up in several economies, including Bangladesh, India, the United Kingdom, the European Union, the United States, Pakistan, Mexico, Indonesia, and the Russian Federation, in 2024, concerns that AI could influence elections are prompting countries to take precautionary measures.

AI can create complex networks to undermine the sanctity of the ballot box, manipulate voter perceptions, and carry out micro-targeted propaganda against election infrastructure, in contrast to the obvious threat of cyberattacks.

During the US Presidential elections, data science company Cambridge Analytica ran a massive advertising campaign to target voters based on their psychology. As a result, voters received different ads based on their perceptions and susceptibility to debates. This was possible due to the availability of real-time data on voters based on their social media behavior, consumption patterns, and relationships. These digital footprints were used to build unique behavioral and psychological profiles of cyber citizens, which could then be targeted directly.

In 2024, the traditional way of conducting elections has completely changed. Earlier elections used to involve campaigning with physical rallies, slogans, and ballot papers. Over time, the ballot papers have changed to electronic voting machines and physical campaigning has changed to social media posts, videos, extensive media coverage, etc.

AI redefines electoral systems:

1. AI-powered bots and tools are used by political parties to: Monitor engagement View social media content and posts in real-time.

2. AI technologies such as speech recognition and natural language processing are being used to make voting more accessible to people with disabilities by allowing them to vote independently using assistive devices. Masu.

3. Predictive analytics powered by AI can predict the potential impact of campaign activities ranging from ad placement to public statements and policy announcements.

4. Election authorities can use AI to monitor and identify violations of election and financial laws and ensure compliance with election rules. In 2021, the Bihar State Election Commission partnered with AI firm Staqu to analyze CCTV footage of counting booths during panchayat elections using optical character recognition (OCR) video analysis. Drones are being used to monitor election events and rallies.

Misuse of AI affects a variety of stakeholders.

The integration of AI into electoral processes raises several important issues that need to be carefully considered.

1. Data Analysis and Voter Targeting – AI analyzes vast amounts of data to identify voter preferences, behavior, and demographics, allowing it to more effectively target specific voter segments. In 2005, popular search engines added programs for targeted advertising that personalize content according to users’ supposed interests. As a result, voters are forced to consciously absorb content that influences their subconscious.

2. The spread of misinformation and deepfakes – The World Economic Forum’s Global Risks Report 2024 ranks misinformation and disinformation as the biggest risk over the next two years. The spread of AI-based misinformation is like a tsunami: powerful, widespread, and literally unstoppable. Recent deepfake videos of Bollywood actors Amir Khan and Ranveer Singh, a video of the president of Moldova, and an audio clone of the US president are examples of how deepfakes interfered with elections.

3. The role of the media in today’s electoral systems cannot be undermined by its wide scope and accessibility to the public.

4. bias and fairness – AI systems can perpetuate and exemplify biases present in the data they are trained on, disadvantaging certain demographic groups and thereby building bias into electoral processes such as voter registration, candidate selection, and even outcome prediction. A classic example of AI perpetuating bias is when Google’s AI platform suggested the policies of the Indian Prime Minister were “fascistic,” leading to a social media discussion about political propaganda to create influenced narratives.

5. Weapons against women – In 2023, a doctored image of a female wrestler smiling broadly while protesting against the president of the wrestling federation (she looked serious in the original photo) went viral, causing the wrestler to lose credibility. In the run-up to elections in Bangladesh, a deepfake video of a female opposition politician wearing inappropriate clothing has emerged, an attempt to target the integrity of a particular socially disadvantaged group for the benefit of another group. This is one example.

6. geofencing – The geotargeting industry gets location data from an online ad placement system known as “real-time bidding,” in which publishers or app owners share their device, ad unit, and geolocation information in a split-second auction. In 2018, a conservative political party reportedly used location data to target people who frequented a Catholic church in Missouri over a 60-day period. This gives political parties and bad actors an additional edge and advantage over voter preferences.

7. Use of Data Without Consent – The use of AI in election campaigns raises ethical concerns about using personal data and behavioral patterns to manipulate voter perceptions through data that is already publicly available, making it difficult to assign legal liability .

8. Foreign interference – Russia was accused of interfering in the 2016 US election, which was believed to have influenced the outcome. The rise and availability of advanced AI-based classification tools has enabled party arms and national security agencies to categorize vast amounts of data into accurate, actionable information down to the individual. This could have fatal consequences for political parties.

9. Honesty – Days before Slovakia’s elections, an audio recording was leaked on Facebook purporting to be a conversation between candidates and media representatives discussing plans to manipulate the election, including vote buying.

1. The deepfake video showed a member of a major political party urging people to vote for his son, who was the party’s candidate in an election. The member died four years ago, but his likeness was recreated using AI tools.

2. The hugely popular DPP in Delhi used “Blaster,” an AI-generated voice of a candidate that greets call recipients by name.

How will deepfakes affect election results?

If false reports have widespread impact, they could completely sway public opinion in the run-up to an election and call into question the democratic landscape of any country.

Response Actions

Social media intermediaries offer the following mitigating measures:

1. meta The company announced it was launching a new helpline in collaboration with the Misinformation Alliance (MCA), which aims to “combat media generated using artificial intelligence”.

2. whatsapp We have introduced a new feature that allows users in India to report suspicious content through an in-app multilingual helpline chatbot. Supports English, Hindi, Tamil and Telugu.

3. Google The company also announced that it will work with various leading technology companies to prevent false AI-generated image, audio, and video content from interfering with global elections.

4. Logical.ai is one of the world’s largest commercial fact-checking organizations, helping to identify false information and content spreading in the public sphere.

5. In 2021, Adobe, Microsoft, Intel, Arm, BBC and Truepic Launched a coalition for standards development. Coalition on Content Origin and Authenticity (C2PA) is a consortium that accelerates the pursuit of practical, adoptable standards for digital origins, publishers, digital creators, media platforms, and consumers.

European Union Artificial Intelligence Act 2024 (EU Law)

EU AI law classifies AI systems into “high risk” categories based on the strength of their impact, lays out transparency requirements to be observed, and imposes an obligation to label AI-generated content.

US

Michigan will become the fifth state after California in November 2023 to regulate AI in election communications and punish those who knowingly disseminate AI-generated deepfakes in election communications. 90 days before the election.

Guidelines issued by the Federal Election Commission

The Federal Election Commission in the United States has issued guidelines for political parties to disseminate public information and requires them to clearly and conspicuously display a disclaimer.

India

1. MEITY plans to amend the Information Technology (Intermediaries Guidelines and Digital Media Ethics Code) Rules 2021 (amended on April 6, 2023) and repeal the following rules: AI Bias,check, Prevent deepfakes And then Regulate fraudulent loan apps On social media.

2. The government also promulgated the Digital Data Protection Act 2024.

3. However, the proposed amendments and the Digital Personal Data Act 2023 will only be notified after the Sabah elections. To fill this gap, MEITY has proposed a new law to deal with deepfakes, where current laws are actionable under the Indian Penal Code, 1860 and the IT Act, 2000 and the Information Technology (Interim Guidelines and Digital Media Ethics). We have issued an advisory stating that we have adequate facilities. Code) Regulations, 2021.

4. Election Commission of India (ECI) We have published a guideline titled. “During the MCC period for general elections and by-elections, social media platforms will be used responsibly and ethically, and abuses by political parties and their representatives will be strictly avoided.” A presidential decree dated May 6, 2024, instructed political parties to remove deepfakes within three hours.

5. The ECI issued a new press note on April 7, 2024, urging young people and first-time voters to take part in the elections through the “Turn 18 Campaign”, while also highlighting the prevalence of fake news and misinformation online.

6. Additionally, a National Cybercrime Reporting Portal will enable cyber citizens (netizens) to report cybercrime cases, including deepfakes.

As the world comes together to address the challenges posed by rapidly developing technology, it is imperative that AI be regulated to ensure the truth and reliability of the information made available to the public. The EU’s enactment of its Artificial Intelligence Act has paved the way for other countries to follow.

About the author: Mr. Vikrant Rana is the Managing Partner of SS Rana & Co. Mr. Anuradha Gandhi is his Managing Associate and Ms. Rachita Thakur is his Associate.

Deals, columns, press releases Bar & Bench, Please fill out the available form here.