Adtech regulation a priority for Modi govt in third term: MoS Rajeev Chandrasekhar

Credibility of news and platforms are as important as monetisation challenges: Anurag Thakur

HUL: Meet the leaders who have emerged from "The CEO Factory"

Punit Goenka addresses Zee employees; ensures no layoffs due to failed merger

Disney's strategic sale to Viacom18 set to boost market share to 40 percent

The Leader's Edge: Storyboard18 in conversation with leaders, decision makers and disruptors

Nobody’s Perfect Retro Porsche Ad

In Conversation With Raja Rajamannar, Mastercard

Shahid Kapoor And Varun Ganjoo Speaks On PokerBaazi's Campaign

In conversation with IHCL’s Puneet Chhatwal & GroupM’s Kate Scott-Dawkins

Why are brands including e-sports in their sports marketing itinerary?

Storyboard18 x The English Nut: How to pronounce these 10 brand names

Experts' Take On Reinvention Of Ad Campaigns Through Technology

Experts Speak On Key Emerging Trends In Cannes Lions Entries

Leaders of the future have to be outstanding human beings: Nestle MD

In conversation with VMLY&R’s Hari Ramanathan

The Indian Computer Emergency Response Team (CERT-In), part of the Ministry of Electronics and Information Technology (MeitY), is currently testing anti-deepfake technology to combat the malicious use of Artificial Intelligence (AI) in scamming unsuspecting individuals. Recently, Sunil Bharti Mittal, Chairman of Bharti Enterprises, shared a shocking account of an AI-driven scam that targeted his company.

At the NVIDIA AI Summit in Mumbai, a source familiar with CERT-In’s efforts confirmed that the agency is actively testing this anti-deepfake technology. “The technology will not only detect deepfakes but will also support prosecution of such offenders in a court of law.”

Referring to the Airtel Chairman’s experience, the source added, “The software can detect fake audio as well as video, which will aid in combating fake news that stirs public alarm. MeitY immediately takes down such videos in collaboration with social media intermediaries.”

One recent case involved SP Oswal, Chairman of Vardhman Group, who lost ₹7 crore after scammers, posing as government officials and using fake documents along with virtual settings, persuaded him to transfer funds.

The Election Commission expressed serious concerns about Deepfake AI and warned of strict action against those spreading fake narratives. Chief Election Commissioner Rajiv Kumar warned of strict measures against those using Artificial Intelligence and deepfakes to spread misinformation about elections. 

In December 2023, CERT-In issued an advisory informing citizens about the rising use of AI and deepfake technology by scammers. According to the advisory, scammers harvest social media profiles and other websites to gather videos, photos, and other personal details, then create convincing deepfakes that mimic real voices and faces of the victims’ friends or family. These deepfakes are used in schemes like fake kidnapping or ransom calls, creating urgency to coerce victims into sending money, often in the form of gift cards or cryptocurrency.

In November 2023, Minister of Information and Broadcasting Ashwini Vaishnaw announced that the government would begin drafting regulations specifically for deepfakes. After consultations with multiple stakeholders and platforms, Vaishnaw stated, “We will start drafting the regulations today, and very soon, we will have specific regulations for deepfakes.”

These forthcoming regulations are anticipated to include penalties for creating or sharing deepfakes and may also establish guidelines to help users identify deepfake content.

Additionally, Google has partnered with the Election Commission of India (ECI) to share crucial voting information via Google Search and YouTube. Google also announced its support of Shakti, The India Election Fact-Checking Collective, and joined the Coalition for Content Provenance and Authenticity (C2PA) as a steering committee member. C2PA is a global standards organization dedicated to certifying the authenticity of digital content.

Meta, in collaboration with the Misinformation Combat Alliance (MCA), launched a dedicated WhatsApp helpline to tackle AI-driven misinformation, particularly deepfakes. Furthering its consumer education efforts, Meta also rolled out initiatives like the ‘Know What’s Real’ campaign on WhatsApp and Instagram to educate users on recognizing and reporting suspicious content.

MeitY's CERT-In trials anti-deepfake tech to combat AI-driven scams

The Delhi High Court has demanded that the Central government disclose any measures taken to address the growing problem of deepfakes. A Division Bench, led by Chief Justice Manmohan and Justice Tushar Rao Gedela, observed the rise of deepfakes, digitally manipulated videos that impersonate individuals and can be used to spread misinformation, as per reports.

Recognizing the gravity of the situation, the Court questioned Additional Solicitor General Chetan Sharma about the government's establishment of an internal mechanism or committee to investigate countermeasures. In the absence of such an initiative, the Chief Justice indicated the Court's intention to form a committee to address this pressing issue.

The Delhi High Court has demanded transparency from the Central government regarding its efforts to combat deepfakes. The Court expressed concern that the public needs to be informed about the measures being taken and the qualifications of any government-established committee. The Additional Solicitor General, representing the Union Ministry of Electronics and Information Technology, assured the Court that the matter is under review and that relevant personnel are in place. However, due to the sensitive nature of the issue, he indicated that full disclosure might not be feasible. The Court, while not requiring the committee's specific details, insisted on information about the progress made. The ASG was instructed to submit a status report within three weeks.

The Court is handling two petitions filed by journalist Rajat Sharma and advocate Chaitanya Rohilla, seeking regulations for deepfakes and measures to identify and block websites disseminating such content. The Court had previously expressed shock at the widespread availability of deepfakes, noting that individuals can no longer trust their own eyes and ears.

Senior Advocate Darpan Wadhwa, representing Rajat Sharma, highlighted the disproportionate impact of deepfakes on women, who are often targeted with sexually explicit content. The Chief Justice acknowledged the challenge of preventing the misuse of artificial intelligence technology while recognizing its potential benefits. The ASG suggested developing countermeasures to mitigate the negative effects of AI.

Wadhwa criticized the government's approach of treating websites disseminating deepfakes as online intermediaries. He argued that the current advisory requiring the removal of deepfake content within 72 hours is insufficient and that immediate action is necessary.

The Chief Justice also referenced the recent hoax bomb threats on Indian airplanes and emphasized the need for enhanced infrastructure to address such tech-enabled threats. The case will be heard again on November 21.


Delhi HC demands deepfake clarity from central government

A day ago, cricketer Virat Kohli became the target of a deepfake video that criticized top-order batsman Shubman Gill. It's important to note that miscreants are not only using AI to spread malicious content but are also using deepfakes to manipulate video KYC (Know Your Customer) procedures of financial institutions in India. 

Sources claim that for the past four months, both central and state law enforcement agencies and Forensic Science Laboratories (FSL) in the country have been testing anti-deepfake softwares not only to investigate crimes but also to detect them early. In July this year, India’s National Stock Exchange (NSE) warned investors about deepfake videos falsely featuring its CEO, Ashishkumar Chauhan, in some investment and stock advisory videos. 

Recently, the Mumbai police filed a case against the Maharashtra Youth Congress’ social media account and 16 others for allegedly sharing a deepfake video of Union Home Minister Amit Shah. Multiple well-known personalities, including Barack Obama, Nancy Pelosi, Elon Musk, Ukrainian President Volodymyr Zelenskyy, Tom Cruise, Scarlett Johansson, and Emma Watson, have also fallen victim to deepfake videos on the internet. Similar such cases have been reported widely across the country.

Sources told Storyboard18 that several states in the country are already using anti-deepfake technology for investigations and to detect the malicious use of AI for fraudulent purposes on the trial basis. It will be announced by the Central government this year.

A source from cybersecurity industry on condition of anonymity said, “State government like Maharashtra is already testing anti-deepfake software on trial basis. The central government will announce about this initiative this year.”  

Ankush Tiwari, Founder and CEO of Pi Labs, said, “We have seen multiple cases of deepfake videos targeting politicians and celebrities going viral on social media platforms. This is clearly mischievous and criminal. But beyond that, miscreants are also using deepfakes to manipulate financial institutions. With central government schemes directly linked to such bank accounts, these miscreants can easily bypass video KYC, causing the original beneficiaries to suffer.” 

Authentify by Finacus and Pi Labs, which helps detect deepfakes and prevent video KYC fraud, is working with multiple clients in the BFSI (Banking, Financial Services, and Insurance) sector. Tiwari explained, “Video KYC is under threat, and that threat comes from AI. So, we are using AI to combat AI’s malicious use. Authentify uses extensive training data of various synthetic content generated by different mechanisms and engines. With this engine, we detect any synthetic content in video and audio files.” 

Tiwari continued, “We’re seeing early signs of these cases occurring in the country. Fraud is being committed using cloned audio and recorded videos. I believe we are not far from witnessing the next Jamtara-like scams in the country. Overall, this is on the rise, and as a society, we need to be prepared. We are on a mission to protect our citizens, enterprises, and government from the malicious use of AI.” 

In November 2023, Minister of Information and Broadcasting Ashwini Vaishnaw announced that the government would begin drafting regulations specifically to address deepfakes. Vaishnaw after meeting multiple stakeholders and platforms had said, "We will start drafting the regulations today, and within a very short timeframe, we will have separate regulations for deepfakes,” Vaishnaw said. 

The new regulations are expected to include penalties for those who upload or create deepfakes and may also establish guidelines to help users distinguish between deepfakes and original content.

Exclusive: Centre, State governments testing anti-deepfake tech, likely to launch this year

With the Paris Olympics on the horizon, cybersecurity firm McAfee has discovered deepfake scams that are designed to discredit the International Olympic Committee (IOC). The scammers have used the deepfake version of billionaire Elon Musk's voice to raise their propoganda this time.

As per CNBCTV18 report, in the latest version, the scammers have created a fictitious Amazon Prime series titled 'Olympics has Fallen II: The End of Thomas Bach', released on Telegram last month. 

The series has targeted the Olympics by spreading misinformation and creating believable narratives. In the series, the scammers have used the deep fake voice of Elon Musk's voice for the narration. 

McAfee said that Musk's voice is a common target for scammers. The company's 2023 Hacker Celebrity Hot List ranked him as one of the most frequently mimicked public figures in cryptocurrency scams.

<a href="https://www.storyboard18.com/how-it-works/youtube-now-allows-users-to-request-removal-of-ai-generated-deepfakes-36090.htm">YouTube now allows users to request removal of AI-generated deepfakes</a>

Last month, a report from Microsoft claimed that Russia was targeting the Paris Olympics with a disinformation campaign that included deploying a deepfake Tom Cruise, narrating a documentary, and criticizing the organization behind the games. 

Notably, Russia has been banned from the 2024 Olympics, although a small number of its athletes may compete as neutrals. 

<a href="https://www.storyboard18.com/how-it-works/when-tech-goes-bad-celebrities-and-the-dangers-of-deepfake-technology-32333.htm">When tech goes bad: Celebrities and the dangers of deepfake technology</a>

The IOC called the nine-minute episodes an "organized disinformation campaign" against it. 

The Olympic body said, "The International Olympic Committee (IOC) has recently been faced with several fake news posts targeting the IOC. An entire documentary produced with defamatory content, a fake narrative, and false information, using an AI-generated voice of a world-renowned Hollywood actor".

McAfee has predicted an increase in scams using AI as the Olympics is around the corner. However, the cybersecurity firm has cautioned people to avoid unfamiliar websites, and unsolicited messages and verify the information source. 

Deepfake video targets Paris Olympics with fake Amazon Prime series using  AI 'Elon Musk' voice

Adobe released findings of its Future of Trust Study for India, which surveyed over 2,000 Indians about their experience encountering misinformation online and concerns about the impact of generative AI. Tied to its work on Responsible Innovation, Adobe commissioned the study to understand public perceptions and sentiment about the societal impact of generative AI and misinformation, including its implications within the context of the ongoing elections in India. Among the findings, the study highlights strong desire from consumers to have tools to verify the trustworthiness of digital content and the urgent need for proactive measures to address misinformation’s potential impact on election integrity in the country.

“We are excited about the potential for generative AI to enhance creativity and productivity, but it is also a transformational technology that demands thoughtful consideration of its societal impact. Our Future of Trust Study underscores the urgent need for media literacy campaigns to educate the consumers about the dangers of deepfakes and to empower them with tools to discern fact from fiction. With the ongoing elections, now is the time to embrace protective technologies like Content Credentials to rebuild trust in the digital ecosystem,” says Prativa Mohapatra, Vice President & Managing Director, Adobe India.

<strong>Key findings of the study include:</strong>

Misinformation is regarded as one of the biggest threats facing society - a concern that may be eroding trust in certain platforms where misinformation is spread.
With misinformation becoming more prevalent, the study shows that 81% of Indians fear that the content they consume online could be altered to fuel misinformation, while an equal percentage also said that it is becoming difficult to verify whether the content, they are consuming online is trustworthy.   

Concerns about misinformation are also impacting consumer’s use of social media platforms with 45% Indians saying that they have stopped or curtailed their use of a specific social media platform due to the amount of misinformation on it.

Amidst a climate of eroding trust in digital content, Indians are concerned about protecting election integrity.

In a year when over four billion people globally are expected to head to the polls, the study found that a significant number of Indians (86%) believe that misinformation and harmful deepfakes will impact future elections.

Without widespread tools to help them decide whether the online content they are consuming is true, a significant number of Indians (82%) believe that election candidates should be prohibited from using generative AI in their promotional content.

Most Indians (94%) believe that governments and technology companies should work together to protect election integrity against the detrimental effects of deepfakes and misinformation.

Most Indians consider the credibility of a source before sharing content online but also believe that having the tools and media literacy skills to verify whether a content is true is important.

Most Indians (92%) believe that it is essential that they have the right tools to verify if online content is trustworthy.

In addition, high percentage of Indians (93%) agree that it is important to know if the content they are consuming is generated using AI.

Concerns regarding misinformation are impacting how people think about its potential effect on children, with 94% of Indians saying that children should be taught media literacy skills as part of their school curriculum.

Indians want govts, tech companies to collaborate to protect election integrity: Adobe Report

The viral video of actress Rashmika Mandanna took the internet by storm, garnering more than 14 million views since being put up online. This raises the question about the legal framework of deepfake cases in India.  Union Minister Rajeev Chandrasekhar responded to a deepfake video of the actor saying that such "damaging form of misinformation need to be dealt with by platforms". 

Deepfakes are the 21st century’s answer to Photoshopping, deepfakes use artificial intelligence called deep learning to create images of fake events. 

Amitabh Bachchan who played the role of Mandanna's father in 'Good Bye' supported the actor and called for a legal action. 

Taking to X, the Union Minister of State (MoS) for IT said that the central government under Prime Minister Narendra Modi is "committed to ensuring [the] safety and trust" of all citizens using the internet.


<blockquote class="twitter-tweet"><p lang="en" dir="ltr">PM <a href="https://twitter.com/narendramodi?ref_src=twsrc%5Etfw">@narendramodi</a> ji&#39;s Govt is committed to ensuring Safety and Trust of all DigitalNagriks using Internet<br><br>Under the IT rules notified in April, 2023 - it is a legal obligation for platforms to<br><br>➡️ensure no misinformation is posted by any user AND<br><br>➡️ensure that when reported by… <a href="https://t.co/IlLlKEOjtd">https://t.co/IlLlKEOjtd</a></p>&mdash; Rajeev Chandrasekhar ?? (@Rajeev_GoI) <a href="https://twitter.com/Rajeev_GoI/status/1721398000701063460?ref_src=twsrc%5Etfw">November 6, 2023</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>

Referring to the IT rules notified in April 2023, the Union Minister referred to the IT rules notified in April 2023 and made it clear that the guidelines that social media platforms must abide by as it a "legal obligation". Chandrasekhar said that all platforms must "ensure no misinformation is posted by any user" and that "when reported by any user of [the] government, misinformation is removed in 36 hours".

He specified that incase social media platforms fail to comply with the rules, then rule 7 will be applied and "platforms can be taken to court by [the] aggrieved person under provisions of [the] IPC".







Rashmika Mandanna deepfake viral video: Union Minister Rajeev Chandrasekhar reacts, says ‘dangerous, damaging’

Hrithik Roshan is a special forces agent in a bulletproof vest who is seated in a black van, holding a gun. Along with him are several other agents. Suddenly, they hear a knock on the door. As soon as they open it, the group point their guns at the intruder - a Zomato delivery man who is as shocked as they.

As one of them asks, "Yeh kisne mangaaya?" Roshan replies, “Maine! Burger ka mann kiya. Sirsa main hain, toh Grillmasters se mangaliya.” But the name of the dish, the restaurant and the city and locality changes every time depending on the ad’s viewer - and their phone’s GPS.

This is Zomato’s hyper-localised campaign that claims to use deepfake technology to deliver personalized messages to audiences.

<iframe title="YouTube video player" src="https://www.youtube.com/embed/_rTEWgQHTcI" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"></iframe>

At the 2022 Cannes Lions International Festival of Creativity, another hyper-localised campaign - ‘Not Just a Cadbury ad’, by Ogilvy India and Wavemaker, which featured actor Shah Rukh Khan, brought home a Titanium Lion. The brand made use of machine learning to recreate the actor’s face and voice to highlight the names of various local stores in the ad, extending help to hundreds of small business owners who were deeply affected by the lockdown.

<iframe title="YouTube video player" src="https://www.youtube.com/embed/5WECsbqAQSk" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"></iframe>

Pepsi and Ageas Life Insurance have also used the technology to create a much younger Salman Khan and Sachin Tendulkar in their commercials.

<iframe title="YouTube video player" src="https://www.youtube.com/embed/lYIj7rkR8VY" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"></iframe>

Brands are playing and experimenting with new technologies to give audiences something new and to break through the clutter.

<strong>But what is the technology and how does it work?</strong>

Rephrase.ai’s founder Ashray Malhotra refers to the technology driving these campaigns as ‘synthetic media’. “Every person thinks synthetic media is deepfake. If someone even generates Shah Rukh Khan’s voice, that gets labelled as deepfake voice. But, that is not how it began,” says Malhotra, whose company uses artificial intelligence technology to create hyper-personalized videos for brands and businesses.

In fact, deepfake tech turned into a menace and continues to be so.

Deepfake technology, which is a combination of two concepts ‘deep learning’ and ‘fake’, was initially used for creating pornographic representations of women. Today, this technology is also used to fool recruiters and voters.

But it has other useful applications. Marketers have begun using the technology to help brands communicate and reach their audiences in what they hope are more engaging ways.

<strong>How it works</strong>

Speaking about the Zomato campaign featuring Hrithik Roshan, the company's spokesperson tells us, “Business owners—big and small—dream of having Bollywood celebrities endorse them. We wanted to give a platform to the restaurant partners that have helped Zomato reach over 1000 cities and have helped serve millions of consumers.”

Four months in the making, the company shares, “AI technology married with some smart targeting on Google and Facebook helped us land a relevant message for the viewers at a mass scale.”

Zomato partnered with Gan Studio, a startup with focus on state of art algorithms.

This campaign was also advertising agency Enormous’ first experience with experimenting with AI-driven technology in crafting a narrative around the brand and its proposition. The agency’s managing partner Ashish Khazanchi mentions that this technology is of great use when the distribution of advertisements is through digital.

“When it comes to distribution, deepfake does make for a lot of interactivity, and a lot of personalisation of assets. It works well when you are trying to create personalized content for people. Therefore, you can have any number of output which would work very well for a hyperlocal market, or a pin pointed kind of a distribution strategy,” explains Khazanchi.

Saumya Rathor, Category Lead, Pepsi Cola, PepsiCo India, agrees with Khazanchi. She says, "With the acceleration of a digital forward communications approach, audiences are turning towards a technologically advanced way of consuming information."

Malhotra looks at it as an evolution of marketing. “The end goal of a marketer is to talk to their audience in a personalized manner, and in a video format which is the most effective. Hence, personalization plus videos is what synthetic media now enables, and one can see more and more marketers make use of this technology,” he says.

Bengaluru-based Rephrase.ai builds generative AI tools to ease video creation. Its clients include Cadbury, Castrol, Raymond, online teaching platform Classplus, edtech startup Sunstone, and learning platform White Hat Jr as clients.

Rephrase.ai has also worked with CXOs like Titan Company’s Ajoy Chawla, CoinDCX’s Minal Thukral, DBS Singapore’s Royce Teo, Becton Dickinson’s Pavan Mocherla, Bagic’s Tapan Singhel, Johnson &amp; Johnson Vision’s Sherri Ferstler, and Meesho’s Vidit Aatrey.

<strong>Inclusion of deepfake in advertising: Critical stages</strong>

Khazanchi explains that the critical stage is always finding the right kind of idea where technology would be of use.

In the case of Zomato, it made sense to the agency due to specific restaurants, and specific dishes in specific parts of the country. He says, “You need to shoot in a static block where you do not need too many changes in the backdrop. You will want the person to be stationed in that particular place. There is enough data captured using a celebrity so that all we have is the information on the way he or she would speak or intonate any particular part of him or her.”

Speaking of the Pepsi campaign, which recreated a scene from the 1994 Bollywood super hit Hum Aapke Hain Kaun, Rathor says, "To start with ensuring that we got the right script to do justice to an iconic character of Indian cinema was a task. We had to ensure that while we make the film quirky, witty, fun, it had to be respectful to the character. At the same time, there is so much love for Prem (Salman Khan’s character in Hum Aapke Hain Kaun) that we knew there would always be high expectations from the audiences while recreating this avatar."

<strong>That's the tale. Now, what about tech?</strong>

Rephrase.ai’s Malhotra recalls the three steps involved while rolling out a campaign involving deepfake technology. Step one: creating a clone. A video is recorded for a few minutes. Then the video is handed over to a machine learning algorithm, which trains for a few days, and creates a clone of the person’s face and voice.

Step two is integration and testing. Malhotra explains, “This process is different for every company. If you want to integrate with your customised relationship management (CRM), which means, as soon as a new customer signs up, they get a personalised video from, say the CEO of the company, we need to integrate the AI generated clone in the middle of the pipeline.”

Step three is deployment, where it is decided that this is a live process. Anybody can type in a name and a video gets generated which can be sent via any medium. “The timeline for the creation of a campaign with the inclusion of this technology on an average takes two weeks. For a concerned campaign, we have done it in as short as five to six days,” adds Malhotra.

For ‘Not Just a Cadbury ad’, it took two weeks and Rephrase.ai created five different avatars of Khan. This was done parallelly because he was in five different locations. Plus, Google Maps API was used to highlight the names of the shops. Malhotra mentions, “The initial intention was to create 20,000 videos. But, we went as high as 1,40,000 videos in a span of 10 days.”

Given Salman Khan's popularity, Pepsi's ad was also well received by audiences and "got social media brimming with conversations through media and memes on the return of Salman's iconic character 'Prem'," adds Rathor.

<strong>Putting in checks and balances</strong>

Since Rephrase.ai works with well-known personalities, it recommends companies add the ‘Assured by Rephrase.ai’ logo. Malhotra says, “The logo assures brands that even if something goes wrong, they can say that the video was AI generated, and not of the actual celebrity or executive.”

In the case of celebrities, every video is created with written consent. Malhotra says, ‘Even though brands have the ability to create new videos, they don’t have the authority until and unless they obtain consent. When the final video—besides the written scripts—is created, they are also sent to the celebrity for approval to ensure that they are comfortable with the final video going out. This means that in every part of the process, they are in control of what goes out.”

In the case of customers, ‘Assured by Rephrase.ai’ logo is added to highlight that the video was created with permission and consent of the person in the video.

<strong>Can’t fake creativity</strong>

For agencies and creators, there’s also the additional challenge of how to keep refreshing the use of the technology in campaigns.

Khazanchi says the challenge will always be creativity, and not technology. “It is not that challenging to shoot because the basics are simple. Now what to do, and who to do it for, that is where the challenge lies,” he adds.

The managing partner says, “The bigger challenge would also be crafting a new narrative so that it always stays fresh. Since three to four brands have already experimented with this technology, you have to find a way of making use of it which makes a lot of sense and is still new.”

The point is to find a way to marry creativity and technology to create clutter-breaking work, and not let technology consume creativity.

DEEPFAKE TECHNOLOGY