(India's new IT Rules amendments mandate clear labeling of AI-generated content to combat deepfakes, requiring platforms and creators to identify synthetic media and protecting citizens from misinformation.)
The
internet used to be a great place for everyone to gain knowledge and share
ideas freely. But now, it’s facing a massive problem: it’s getting hard to tell
the difference between what’s real and what’s fake. The recent media reports reveal the fake
social media videos in the Bihar Elections.
Election Commission has issued a warning to the creators of such
deepfake content. This is a great risk
that innocent citizens are facing now.
The
Government of India has now suggested important changes to the IT Rules,
2021, specifically to deal with synthetically generated information—content
made by Artificial Intelligence (AI). These changes are not just small
technical updates; they are a necessary and urgent step to protect every
citizen from a global digital threat.
The
Deepfake Danger: A Crisis That Has Reached India
AI tools
have made it incredibly easy to create "deepfakes"—fake videos,
audios, and images that look and sound completely real. What once took a
Hollywood studio months to do can now be done instantly on a basic computer.
Around the
world, deepfakes are being used for terrible things. They are used to create
fake private images, spread political lies, and even trick companies out of
large sums of money. In India, there are reports about communal tensions based
on these fake videos. The fraud is
getting so advanced that criminals can use AI to fake a person's face or voice
to pass security checks, a process known as biometric spoofing.
In India,
this threat is very real. We saw how quickly deepfake videos of famous actors
and politicians went viral, creating confusion and damage to their reputations.
These fake contents, including videos and audios, are designed to cause social
friction, influence elections, and make people doubt real news. The sheer
volume of this fake content is growing rapidly.
Why We
Must Act Now
Because
AI-generated fake content is so fast and so realistic, we urgently need a
proper system to keep track of it. Our old methods of content checking—relying
on people or simple software filters—are simply not good enough anymore. It is
almost impossible for a normal person to spot a high-quality deepfake.
The
platforms that host and create this content hold the real power. They must now
take on a major and fair share of the responsibility. AI isn't just a simple
tool; it’s a powerful force that can cause both good and bad. If we allow these
platforms to act without rules, simply claiming they are neutral delivery
services, we invite disorder and fail in the basic duty of protecting our
citizens.
How India
Tried to Fight Fake Content Before
India has
always had laws to deal with online harm, but they were not designed for the AI
age. Before the IT Rules, 2021, we mainly relied on parts of the IT Act,
2000, and the Indian Penal Code (IPC). These laws covered crimes
like online cheating or spreading indecent material, but they mostly dealt with
the result of the crime, not the technology behind it.
Older
rules, like the IT (Intermediaries Guidelines) Rules, 2011, told
platforms to warn users about prohibited content but lacked the power to handle
clever AI tricks. The later amendments to the 2021 Rules helped by making
platforms act quickly to remove illegal content. However, they still lacked
clear rules for the specific threat of synthetic content. These new draft rules
are intended to close that critical gap.
New Rules,
Clear Responsibilities
The
current amendments create a very clear system of shared responsibility. The new
legal definition of 'synthetically generated information' ensures that
content that looks real but is AI-made is treated specially.
What
Generative Platforms Must Do (Rule 3(3)):
Any
company that provides tools to create or change synthetic content must embed a
permanent and unique digital marker (metadata) on that content. This marker
must be clearly visible, covering at least 10% of the screen area of a video or
during the first 10% of an audio clip. This ensures that the fake nature is
easy to see and cannot be removed later.
What Big
Social Media Platforms (SSMIs) Must Do (Rule 4(1A)):
Platforms
with a large number of users must follow even stricter rules:
1. Ask the
User:
They must ask the user to confirm if the content they are uploading is
AI-generated.
2. Check the
Claim:
They must use "reasonable technical measures" (including automated
software) to check if the user is telling the truth.
3. Put a
Clear Label: If the content is confirmed as synthetic, the platform
must put a clear and prominent label on it.
If these
platforms fail to follow these rules, they risk losing the legal protection
they currently have. This means that following due diligence is mandatory—it is
the essential requirement for doing business on the Indian internet. This
two-part approach—labeling by the creators and verification by the hosts—is a
model that matches the best new laws being created globally, such as in the EU
and China.
Freedom
with Responsibility
Some
people often argue that strong rules go against the Constitutional right to
freedom of speech and expression (Article 19(1)(a)). However, our
Constitution makes it clear that this freedom is not absolute. It comes
with reasonable limits for the sake of India’s sovereignty, security, public
order, and, most importantly, to prevent defamation and anti-national content.
Freedom of
expression does not give anyone the right to create fake content, lie about
someone, or insult them. Such actions are not free speech; they are digital
attacks. If we allow bad actors to damage people's lives and reputations using
powerful technology, we harm the genuine people who use the internet
responsibly. Accountable freedom, where rights are balanced with duties,
is the only way to keep the internet healthy and useful.
We Welcome
the Change, But Enforcement is Key
PreSense fully
supports this forward-thinking amendment. It sets a necessary and high standard
for accountability in the age of AI.
However, a
great law is useless if it is not used. We urge the Government to make sure
this law is properly enforced. The system of accountability—the Grievance
Officers and the Grievance Appellate Committees (GACs)—must be strong,
well-funded, and able to act fast.
To help
make this work better, we suggest two things:
1. A Simple,
Standard Label: The Government should bring together tech experts to
agree on a single, clear, and simple digital icon or notice that must be
used across all platforms. If every platform uses the same label, people will
quickly learn what the 10% visibility label means, and they won't be
confused by different warnings.
2. Public
Awareness Drive: The rules can only work if users know about them. A
large, public campaign must be started right away to teach the average Indian,
in all major languages, how to spot the synthetic content label and how to file
a complaint if they find something malicious.
3. Setting up
Expert Committee: The
Indian Government should set up an expert committee of representatives from the
Media and cybersecurity organisations and periodically review the
implementation of various rules, through Zoom or any other Indian platform.
4. A system must
be evolved to monitor whether these rules are misused, and implemented only for
the good of society. The latest
amendment does not infringe upon the Fundamental Rights and privacy of people.
It is not against Fundamental Rights, as argued by a few.
This
amendment is a brave and crucial step to ensure that India's digital future is
one of truth and trust, not lies and confusion. We must ensure that the law
stays ahead of the technology.
Here it is
pertinent to remember what a US Supreme Court Justice Oliver Wendell Holmes Jr.
said decades ago: “Your Liberty to Swing Your Fist Ends Just Where My Nose
Begins.” (in modern usage, it is changed as `Your Liberty ends where my nose
begins’ – mostly attributed to Abraham Lincoln) which means Freedom of speech
is essential but that freedom to act ends when it infringes on the rights,
safety or personal domain of other persons. Those who argue for unfettered
freedom of speech should do well to remember the quote.
The
notification may be accessed from the website of Government of India:
https://www.meity.gov.in/static/uploads/2025/10/38be31bac9d39bbe22f24fc42442d5d1.pdf
The official press release
https://www.pib.gov.in/PressReleasePage.aspx?PRID=2181719
Source: This is the Editorial published in October 2025 edition of PreSense
0 Comments