Latest Topics in News
Deepfakes and Democracy: Tackling the Emerging Threat of AI-Generated Misinformation
Deepfakes Recent Incidents and Concerns
A viral deepfake video of an Indian celebrity has sparked controversy in India. The video shows the celebrity’s face morphed onto a woman entering a lift wearing revealing clothes. This incident has raised concerns over online safety and privacy violations, especially for women.
In the lead up to the 2024 Indian elections, multiple political deepfake videos have emerged on the messaging platform WhatsApp. These videos imitate political leaders to spread false information and inflame tensions between parties.
Proliferation on WhatsApp makes it difficult to curb the spread of such malicious deepfakes. WhatsApp’s encryption makes tracking originators challenging.
The unconstrained circulation of political deepfakes raises grave concerns over their ability to manipulate election outcomes in the world’s largest democracy.
The recent incidents highlight the following problems caused by deepfakes:
- Spreading misinformation
- Violating privacy
- Manipulating elections
- Facilitating harassment of women
There is an urgent need for measures to mitigate the risks posed by the proliferation of deepfakes in India. Here is an UCN team’s comprehensive analysis of the Deepfakes, to deliver to you the key concepts and insights needed.
Deepfakes Overview
| Sections | Details |
|---|---|
| What are Deepfakes? |
|
| How are Deepfakes Created? |
|
| Risks and Challenges of Deepfakes |
|
| Deepfakes in India |
|
| Global Policy Approaches |
|
| Tackling Deepfakes |
|
| Role of Ethical Technology Use |
|
| Balancing Innovation and Ethics |
|
Government of India tackling Deepfakes
The Government of India issued a new advisory in Nov 2023, mandating social media platforms, to take down deepfake content within 24 hours of receiving a complaint, as part of the 2021 Information Technology Rules.
Existing Laws Relevant to Deepfakes in India
| Law | Description | Penalty |
|---|---|---|
| Section 66E of IT Act | Applies to deepfake offenses involving recording, publishing, or sending a person’s photographs in mass media, infringing on their privacy. | Up to 3 years in prison or a fine of up to ₹2 lakh |
| Section 66D of IT Act | Allows prosecution for the criminal use of communication devices or computer resources with the intent to cheat or impersonate. | Up to 3 years in prison and/or a fine of up to ₹1 lakh |
Copyright Protection and Deepfakes in India
| Law | Description | Penalty |
|---|---|---|
| Indian Copyright Act, 1957 | Protects works like films, music, and creative content. Copyright holders can sue for deepfakes using copyrighted works without permission. | Penalties outlined in Section 51 of the Copyright Act |
Introduction to Deepfakes
- What are deepfakes?
- How are they created using AI and deep learning?
- Examples of deepfake videos
Deepfakes refer to media content such as videos, images or audio that have been manipulated using artificial intelligence technologies. Specifically, deepfakes are generated using deep learning techniques, a sophisticated form of machine learning. Deep learning algorithms are trained on large datasets of images, videos and audio clips to analyze and recreate intricate patterns. The outputs are often strikingly realistic and difficult to distinguish from genuine media.
Deepfakes involve altering or replacing the likeness of a person through facial mapping and voice imitation. For instance, deepfake technology can stitch a person’s face onto another body or make them appear to say things they never actually said. The seamless doctoring of media made possible by deep learning represents an evolution from rudimentary editing techniques like photoshopping.
High profile deepfake videos feature public figures, realistically edited into fictional situations. Such convincing forgeries demonstrate the power of deepfakes to spread misinformation and erode public trust.
The Power and Risks of Deepfakes
- Ability to manipulate perceptions and spread misinformation
- Undermine trust in institutions and information
- Used to create revenge porn, hack facial recognition
- Blur lines between fact and fiction
Deepfakes possess alarming power to manipulate perceptions and propagate misinformation. Their ability to realistically impersonate people grants them credibility to sway opinions and beliefs. Deepfakes can be employed to malign political candidates by depicting them making offensive comments. Such false narratives undermine trust in public figures and institutions.
Deepfakes also enable unethical violations of privacy. They have been misused to create non-consensual intimate imagery in acts of revenge porn. Deepfakes designed to mimic biometric data can potentially bypass security systems based on facial recognition or voice recognition.
The most disconcerting impact of deepfakes is the blurring of truth and fiction. As technological capabilities improve, deepfakes exhibit fewer imperfections that can expose them as fraudulent. The increasingly plausible forgeries make it difficult to discern what is real, creating confusion and polarization.
Global Approaches to Deepfakes
- EU’s code of practice on disinformation
- US Deepfake Task Force Act
- China’s new regulations on deep synthesis
The European Union updated its Code of Practice on Disinformation in 2022 to combat online manipulation including deepfakes. The code promotes transparent labelling and greater accountability for platforms hosting synthetic media.
In the United States, the proposed Deepfake Task Force Act seeks to empower the Department of Homeland Security to monitor and counter unknown deepfake threats. It signals a recognition of the need for specific governmental initiatives to address deepfakes.
China instituted comprehensive regulations on deep synthesis technology effective January 2023. The strict rules mandate consent requirements, content review systems and close coordination with Chinese authorities to limit disinformation. However, critics argue the regulations may increase censorship and state surveillance.
Tackling Deepfakes: The Way Forward
- Developing AI tools for detection
- Blockchain-based verification
- Policies for deepfake impact mitigation
- Deepfake Accountability Act
- Public awareness campaigns
Technical and policy interventions are required to address the proliferation of deepfakes. AI-powered algorithms can be created to automatically detect manipulated media based on analyzing abnormalities. Social media platforms need to be proactive in deploying such automated fact-checking tools.
Blockchain technology presents a promising solution for deepfake verification. Blockchains can offer tamper-proof records of a media asset’s origin and modification trail. This transparency allows confirmation of authenticity and attribution.
Dedicated policies like the proposed Deepfake Accountability Act can define consequences for maliciously generating or sharing synthetic media. Legal deterrence and impact mitigation frameworks are necessary to curb exploitation.
Mass awareness campaigns by government and media bodies are equally important. Educating citizens about deepfake risks fosters a more informed society and polity. Grassroots advocacy efforts by public-spirited individuals provide foundations for an ethical and responsible deepfake future.
Mitigating Threats: Role of Individuals
- Importance of media literacy
- Ethical technology use
- Responsible sharing on social media
Individuals have a crucial part to play in mitigating deepfake risks by developing media literacy competencies. Learning to critically analyze and question the veracity of content is vital today. Seeking trusted sources to verify information can counter disinformation.
Being judicious in sharing content and reporting suspicious media also helps contain virality. Moreover, refusing to engage in unethical deepfake creation upholds moral values and deters misuse. The aggregate of individual decisions and actions shapes the ultimate technology impact.
Final Thoughts
- Balancing innovation and ethical use
- Collaborative approach needed
In the UCN team’s opinion, Deepfakes exemplify the complex interplay between innovation and ethics. Their capacity for misuse alongside contributions to creative expression highlight the need for nuanced policies. Standards balancing public interest with technological progress require collaborative efforts between stakeholders. With informed debates and collective diligence, deepfakes can be steered towards ethical applications that further human empowerment.
