#FactCheck - Viral Video of Man ‘Running on Water’ Found to Be AI-Generated

The 2020s mark the emergence of deepfakes in general media discourse. The rise in deepfake technology is defined by a very simple yet concerning fact: it is now possible to create perfect imitations of anyone using AI tools that can create audio in any person's voice and generate realistic images and videos of almost anyone doing pretty much anything. The proliferation of deepfake content in the media poses great challenges to the functioning of democracies. especially as such materials can deprive the public of the accurate information it needs to make informed decisions in elections. Deepfakes are created using AI, which combines different technologies to produce synthetic content.

Understanding Deepfakes 

Deepfakes are synthetically generated content created using artificial intelligence (AI). This technology works on an advanced algorithm that creates hyper-realistic videos by using a person’s face, voice or likeness utilising techniques such as machine learning. The utilisation and progression of deepfake technology holds vast potential, both benign and malicious. 

An example is when the NGO Malaria No More which had used deepfake technology in 2019 to sync David Beckham’s lip movements with different voices in nine languages, amplified its anti-malaria message.

Deepfakes have a dark side too. They have been used to spread false information, manipulate public opinion, and damage reputations. They can harm mental health and have significant social impacts. The ease of creating deepfakes makes it difficult to verify media authenticity, eroding trust in journalism and creating confusion about what is true and what is not. Their potential to cause harm has made it necessary to consider legal and regulatory approaches.

India’s Legal Landscape Surrounding Deepfakes

India presently lacks a specific law dealing with deepfakes, but the existing legal provisions offer some safeguards against mischief caused. 

• Deepfakes created with the intent of spreading misinformation or damaging someone’s reputation can be prosecuted under the Bharatiya Nyaya Sanhita of 2023. It deals with the consequences of such acts under Section 356, governing defamation law.
• The Information Technology Act of 2000, the primary law that regulates Indian cyberspace. Any unauthorised disclosure of personal information which is used to create deepfakes for harassment or voyeurism is a violation of the act.
• The unauthorised use of a person's likeness in a deepfake can become a violation of their intellectual property rights and lead to copyright infringement.
• India’s privacy law, the Digital Personal Data Protection Act, regulates and limits the misuse of personal data. It has the potential to address deepfakes by ensuring that individuals’ likenesses are not used without their consent in digital contexts.

India, at present, needs legislation that can specifically address the challenges deepfakes pose. The proposed legislation, aptly titled, ‘the Digital India Act’ aims to tackle various digital issues, including the misuse of deepfake technology and the spread of misinformation. Additionally, states like Maharashtra have proposed laws targeting deepfakes used for defamation or fraud, highlighting growing concerns about their impact on the digital landscape.

Policy Approaches to Regulation of Deepfakes

• Criminalising and penalising the making, creation and distribution of harmful deepfakes as illegal will act as a deterrent.
• There should be a process that mandates the disclosures for synthetic media. This would be to inform viewers that the content has been created using AI. 
• Encouraging tech companies to implement stricter policies on deepfake content moderation can enhance accountability and reduce harmful misinformation.
• The public’s understanding of deepfakes should be promoted. Especially, via awareness campaigns that will empower citizens to critically evaluate digital content and make informed decisions.

Deepfake, Global Overview

There has been an increase in the momentum to regulate deepfakes globally. In October 2023, US President Biden signed an executive order on AI risks instructing the US Commerce Department to form labelling standards for AI-generated content. California and Texas have passed laws against the dangerous distribution of deepfake images that affect electoral contexts and Virginia has targeted a law on the non-consensual distribution of deepfake pornography. 

China promulgated regulations requiring explicit marking of doctored content. The European Union has tightened its Code of Practice on Disinformation by requiring social media to flag deepfakes, otherwise they risk facing hefty fines and proposed transparency mandates under the EU AI Act. These measures highlight a global recognition of the risks that deepfakes pose and the need for a robust regulatory framework.

Conclusion 

With deepfakes being a significant source of risk to trust and democratic processes, a multi-pronged approach to regulation is in order. From enshrining measures against deepfake technology in specific laws and penalising the same, mandating transparency and enabling public awareness, the legislators have a challenge ahead of them. National and international efforts have highlighted the urgent need for a comprehensive framework to enable measures to curb the misuse and also promote responsible innovation. Cooperation during these trying times will be important to shield truth and integrity in the digital age. 

References

• https://digitalcommons.usf.edu/cgi/viewcontent.cgi?article=2245&context=jss
• https://www.thehindu.com/news/national/regulating-deepfakes-generative-ai-in-india-explained/article67591640.ece
• https://www.brennancenter.org/our-work/research-reports/regulating-ai-deepfakes-and-synthetic-media-political-arena
• https://www.responsible.ai/a-look-at-global-deepfake-regulation-approaches/
• https://thesecretariat.in/article/wake-up-call-for-law-making-on-deepfakes-and-misinformation 

‍

Introduction 

Public infrastructure has traditionally served as the framework for civilisation, transporting people, money, and ideas across time and space, from the iron veins of transcontinental railroads to the unseen arteries of the internet. In democracies where free markets and public infrastructure co-exist, this framework has not only facilitated but also accelerated progress. Digital Public Infrastructure (DPI), which powers inclusiveness, fosters innovation, and changes citizens from passive recipients to active participants in the digital age, is emerging as the new civic backbone as we move away from highways and towards high-speed data. 

DPI makes it possible for innovation at the margins and for inclusion at scale by providing open-source, interoperable platforms for identities, payments, and data exchange. Examples of how the Global South is evolving from a passive consumer of technology to a creator of globally replicable governance models are India’s Aadhaar (digital identification), UPI (real-time payments), and DigiLocker (data empowerment). As the ‘digital commons’ emerges, DPI does more than simply link users; it also empowers citizens, eliminates inefficiencies from the past, and reimagines the creation and distribution of public value in the digital era. 

Securing the Digital Infrastructure: A Contemporary Imperative

As humans, we are already the inhabitants of the future, we stand at the temporal threshold for reform. Digital Infrastructure is no longer just a public good. It’s now a strategic asset, akin to oil pipelines in the 20th century. India is recognised globally for the introduction of “India Stack”, through which the face of digital payments has also been changed. The economic value contributed by DPIs to India’s GDP is predicted to reach 2.9-4.2 percent by 2030, having already reached 0.9% in 2022. Its role in India’s economic development is partly responsible for its success; among emerging market economies, it helped propel India to the top of the revenue administrations’ digitalisation index. The other portion has to do with how India’s social service delivery has changed across the board. By enabling digital and financial inclusion, it has increased access to education (DIKSHA) and is presently being developed to offer agricultural (VISTAAR) and digital health (ABDM) services. 

Securing the Foundations: Emerging Threats to Digital Public Infrastructure

The rising prominence of DPI is not without its risks, as adversarial forces are developing with comparable sophistication. The core underpinnings of public digital systems are the target of a new generation of cyber threats, ranging from hostile state actors to cybercriminal syndicates. The threats pose a great risk to the consistent development endeavours of the government. To elucidate, targeted attacks on Biometric databases, AI-based Misinformation and Psychological Warfare, Payment System Hacks, State-sponsored malware, cross-border phishing campaigns, surveillance spyware and Sovereign Malware are modern-day examples of cyber threats. 

To secure DPI, a radical rethink beyond encryption methods and perimeter firewalls is needed. It requires an understanding of cybersecurity that is systemic, ethical, and geopolitical. Democracy, inclusivity, and national integrity are all at risk from DPI. To preserve the confidence and promise of digital public infrastructure, policy frameworks must change from fragmented responses to coordinated, proactive and people-centred cyber defence policies.

CyberPeace Recommendations

Powering Progress, Ignoring Protection: A Precarious Path

The Indian government is aware that cyberattacks are becoming more frequent and sophisticated in the nation. To address the nation’s cybersecurity issues, the government has implemented a number of legislative, technical, and administrative policy initiatives. While the initiatives are commendable, there are a few Non-Negotiables that need to be in place for effective protection: 

• DPIs must be declared Critical Information Infrastructure. In accordance with the IT Act, 2000, the DPI (Aadhaar, UPI, DigiLocker, Account Aggregator, CoWIN, and ONDC) must be designated as Critical Information Infrastructure (CII) and be supervised by the NCIIPC, just like the banking, energy, and telecom industries. Give NCIIPC the authority to publish required security guidelines, carry out audits, and enforce adherence to the DPI stack, including incident response protocols tailored to each DPI. 
• To solidify security, data sovereignty, and cyber responsibility, India should spearhead global efforts to create a Global DPI  Cyber Compact through the “One Future Alliance” and the G20. To ensure interoperable cybersecurity frameworks for international DPI projects, promote open standards, cross-border collaboration on threat intelligence, and uniform incident reporting guidelines. 
• Establish a DPI Threat Index to monitor vulnerabilities, including phishing attacks, efforts at biometric breaches, sovereign malware footprints, spikes in AI misinformation, and patterns in payment fraud. Create daily or weekly risk dashboards by integrating data from state CERTs, RBI, UIDAI, CERT-In, and NPCI. Use machine learning (ML) driven detection systems. 
• Make explainability audits necessary for AI/ML systems used throughout DPI to make sure that the decision-making process is open, impartial, and subject to scrutiny (e.g., welfare algorithms, credit scoring). Use the recently established IndiaAI Safety Institute in line with India’s AI mission to conduct AI audits, establish explanatory standards, and create sector-specific compliance guidelines. 

References

• https://orfamerica.org/newresearch/dpi-catalyst-private-sector-innovation?utm_source=chatgpt.com 
• https://www.institutmontaigne.org/en/expressions/indias-digital-public-infrastructure-success-story-world 
• https://www.pib.gov.in/PressReleasePage.aspx?PRID=2116341 
• https://www.pib.gov.in/PressReleaseIframePage.aspx?PRID=2033389 
• https://www.governancenow.com/news/regular-story/dpi-must-ensure-data-privacy-cyber-security-citizenfirst-approach 

‍

Introduction

In the dynamic intersection of pop culture and technology, an unexpected drama unfolded in the virtual world, where the iconic Taylor Swift account has been temporarily blocked on X . The incident sent a shockwave through the online community, sparking debates and speculation about the misuse of deepfake technology.

Taylor Swift's searches on social media platform X have been restored after a temporary blockage was lifted following outrage over her explicit AI images. The social media site, formerly known as Twitter, temporarily restricted searches for Taylor Swift as a temporary measure to address a flood of AI-generated deepfake images that went viral across X and other platforms. 

X has mentioned it is actively removing the images and taking appropriate actions against the accounts responsible for spreading them. While Swift has not spoken publicly about the fake images, a report stated that her team is "considering legal action" against the site which published the AI-generated images.

The Social Media Frenzy

As news of temporary blockages spread like wildfire across social media platforms, users engaged in a frenzy of reactions. The fake picture was re-shared 24,000 times, with tens of thousands of users liking the post. This engagement supercharged the deepfake image of Taylor Swift, and by the time the moderators woke up, it was too late. Hundreds of accounts began reposting it, which started an online trend. Taylor Swift's AI video reached an even larger audience. The source of the photograph wasn't even known to begin with. The revelations are causing outrage. American lawmakers from across party lines have spoken. One of them said they were astounded, while another said they were shocked. 

AI Deepfake Controversy

The deepfake controversy is not new. There are lot of cases such as Rashmika Mandana, Sachin Tendulkar, and now Taylor Swift have been the victims of such misuse of Deepfake technology. The world is facing a concern about the misuse of AI or deepfake technology. With no proactive measures in place, this threat will only worsen affecting privacy concerns for individuals. This incident has opened a debate among users and industry experts on the ethical use of AI in the digital age and its privacy concerns.

Why has the Incident raised privacy concerns?

The emergence of Taylor Swift's deepfake has raised privacy concerns for several reasons.

1. Misuse of Personal Imagery:  Deepfake uses AI and its algorithms to superimpose one person’s face onto another person’s body, the algorithms are processed again and again till the desired results are obtained. In the case of celebrities or higher-position people, it's very easy for crooks to get images and generate a deepfake.  In the case of Taylor Swift, her images are misused. The misuse of Images can have serious consequences for an individual's reputation and privacy.

2. False narrative and Manipulation:  Deepfake opens the door for public reaction and spreads false narratives, causing harm to reputation, and affecting personal and professional life. Such false narratives through deepfakes may influence public opinion and damage reputation making it challenging for the person to control it. 

3. Invasion of Privacy: Creating a deepfake involves gathering a significant amount of information about their targets without their consent. The use of such personal information for the creation of AI-generated content without permission raises serious privacy concerns. 

4. Difficulty in differentiation: Advanced Deepfake technology makes it difficult for people to differentiate between genuine and manipulated content. 

5. Potential for Exploitation: Deepfake could be exploited for financial gain or malicious motives of the cyber crooks. These videos do harm the reputation, damage the brand name, and partnerships, and even hamper the integrity of the digital platform upon which the content is posted, they also raise questions about the platform’s policy or should we say against the zero-tolerance policy on posting the non-consensual nude images.

Is there any law that could safeguard Internet users?

Legislation concerning deepfakes differs by nation and often spans from demanding disclosure of deepfakes to forbidding harmful or destructive material. Speaking about various countries, the USA including its 10 states like California, Texas, and Illinois have passed criminal legislation prohibiting deepfake. Lawmakers are advocating for comparable federal statutes. A Democrat from New York has presented legislation requiring producers to digitally watermark deepfake content. The United States does not criminalise such deepfakes but does have state and federal laws addressing privacy, fraud, and harassment.

In 2019, China enacted legislation requiring the disclosure of deepfake usage in films and media. Sharing deepfake pornography became outlawed in the United Kingdom in 2023 as part of the Online Safety Act.

To avoid abuse, South Korea implemented legislation in 2020 criminalising the dissemination of deepfakes that endanger the public interest, carrying penalties of up to five years in jail or fines of up to 50 million won ($43,000).

In 2023, the Indian government issued an advisory to social media & internet companies to protect against deepfakes that violate India'sinformation technology laws. India is on its way to coming up with dedicated legislation to deal with this subject.

Looking at the present situation and considering the bigger picture, the world urgently needs strong legislation to combat the misuse of deepfake technology.

Lesson learned

The recent blockage of Taylor Swift's searches on Elon Musk's X has sparked debates on responsible technology use, privacy protection, and the symbiotic relationship between celebrities and the digital era. The incident highlights the importance of constant attention, ethical concerns, and the potential dangers of AI in the digital landscape. Despite challenges, the digital world offers opportunities for growth and learning.

Conclusion

Such deepfake incidents highlight privacy concerns and necessitate a combination of technological solutions, legal frameworks, and public awareness to safeguard privacy and dignity in the digital world as technology becomes more complex.

References:

• https://www.hindustantimes.com/world-news/us-news/taylor-swift-searches-restored-on-elon-musks-x-after-brief-blockage-over-ai-deepfakes-101706630104607.html
• https://readwrite.com/x-blocks-taylor-swift-searches-as-explicit-deepfakes-of-singer-go-viral/

‍