HDFC Bank has rolled out a campaign ‘End of Scam Sale’, to raise awareness about online fraud.
Recently, Deepfake and AI manipulation have become deceptive tools exploiting consumer trust, prompting an urgent need for awareness. After the unauthorised use of actor Rashmika Mandana's face in a deepfake incident in November 2023, the Indian government too issued an advisory as a precautionary measure.
Hence, the brand through its campaign exposes the insidious scams, urging vigilance in the digital age.
Conceptualised by FCB Kinnect, the campaign used deepfake technology, transforming their anti-fraud influencer, Vigil Aunty (Anuradha Menon or popularly known as Lola Kutty), into actor Nora Fatehi. The film showcased Vigil Aunty posed as Fatehi for a fictional brand, ‘Lulumelon’, offering incredible discounts.
Furthermore, the campaign included creating a fake Instagram page and running targeted ads. Users excited about the discounts were directed to the brand's landing page, where they discovered the scam, highlighting the harsh reality of online fraud.
For this campaign, the agency used AI and machine learning to deliver a powerful message about staying vigilant in the face of scams. The final leg of the campaign unveiled the brand's educational initiative, designed to enlighten the audience about the repercussions of fraud.
Campaign India, spoke to Kartikeya Tiwari, senior vice president, and national creative director, FCB Kinnect, to learn more about the campaign.
What inspired the creative vision of HDFC Bank's 'EOSS: End of Scam Sale' campaign, and what motivated the use of deepfake technology and Nora Fatehi's persona?
Two years ago, we introduced a character named Vigil Aunty. As part of our monthly activities, we diligently monitor the prevalent opportunities where individuals may be susceptible to scams or the latest methods employed by scammers. We stay abreast of news and track the evolving strategies employed by scam artists.
Our goal is to inform and raise awareness among the public promptly. While some aspects of the information are real-time, many are predetermined and known to us in advance.
For example, each year, our efforts for the brand are to contribute on a larger scale than the previous year. Hence, when we began planning, concerns arose, especially considering the significant consumer spending during festive seasons, such as Diwali, and sale periods. End-of-season sales, offering substantial discounts like flat 60% or 80% off, become prime scam targets. As a society, we are inclined to embrace such discounts, making us vulnerable to scams.
In the past eight months, we have been developing a campaign, particularly in response to the increasing use of AI in scams. Voice manipulation, exemplified by scammers pretending to be family members in distress, has become prevalent. This realisation prompted us to take action during the peak sales periods to educate people and prevent them from falling prey to these scams.
Our campaign aims to help individuals remain vigilant and discerning, even when presented with seemingly legitimate opportunities, international brands, or endorsements by celebrities.
AI's capabilities make it easier for scammers to orchestrate convincing scenarios, and our initiative seeks to equip people to recognise and avoid such deceptive tactics. This planning exercise led to the formation of our campaign, aligning with the end-of-season sales period. This narrative encapsulates the essence of our endeavour to safeguard individuals from falling victim to scams.
Could you elaborate on the strategies employed to create and promote the fictional fashion brand ‘Lulumelon,’ including the use of AI and machine learning for morphing Fatehi's face on Vigil Aunty?
A seemingly simple thing like our ambition to simulate a scam operation required us to mimic the strategies employed by fraudsters. Our creative strategy closely mirrored that of a group of scammers, and our media strategy was designed in a manner consistent with how scammers plan their media approach.
Creating a strategy involved fabricating a fake brand. We deliberately chose a name resembling that of another renowned brand, mirroring the tactic scammers use, such as adding an ‘e’ to ‘Amazon’. In our case, we took the well-known global brand ‘Lululemon’ and coined ‘Lulumelon’.
We went further by establishing multiple social media handles and various website domains like lulumelon.in, lulumelon.sk, and more. This multi-domain approach aimed to mimic the diversified online presence typical of scams.
To appear authentic as a fashion brand, we needed a recognisable face, someone with genuine fashion credentials. Similar to scammers exploiting the allure of a celebrity, we decided to collaborate with a popular and fashion-forward celebrity.
Unlike scammers, we sought Nora Fatehi’s consent and successfully pitched the idea. She embraced the concept and became the face of our brand, adding a modern and contemporary touch.
The creation of this celebrity's virtual counterpart involved a three-stage process with the assistance of AI partners. First, we meticulously adjusted facial features, proportions, and hairline to make the virtual face resemble the celebrity accurately. Next, we ensured proportionality to the body, refining details for a realistic appearance. Lastly, we seamlessly incorporated the celebrity's voice into our AI model, transforming our main protagonist, Vigil Aunty, into a flawless replica.
This process, while time-consuming, took approximately three weeks to achieve flawlessness. The training time included modelling the voice and perfecting various angles to ensure a convincing transformation. Despite its potential misuse, we harnessed this technology to create a captivating and realistic prank.
Our approach involved creating multiple websites, selecting a celebrity, and utilising AI technology to make Vigil Aunty closely resemble and sound like the chosen celebrity, forming the foundation of this elaborate prank.
Given the potential risks associated with creating a fake brand and using deceptive marketing practices, what ethical considerations were taken into account during the planning and execution of the campaign?
Keeping our intentions above everything is crucial, as that was our primary goal. However, strategically, our planning exercise revealed that the biggest challenge is this thing called ‘optimism bias’.
Today, customers may subliminally think, ‘scams are for others, not me,’ especially if they consider themselves digitally savvy. However, statistics show that the most digitally literate, young people in the country are the ones getting scammed the most. It's not the old or unsavvy.
Why is this happening? Because we are the most vulnerable; we think it can’t happen to us. This is called optimism bias.
Our ethical intention was to give people an experience that is not very shocking, but mildly shocking. We don't want people to get extremely worried. Second, we never even allow people to even possibly transact with real money. When they reach the landing page, and click on one of our ads, within five to 10 seconds, the reveal video is showcased.
Users are unable to add items to the cart or proceed with transactions, given the absence of a payment gateway. In this way, we've ensured that, apart from their time, no additional information or resources are taken.
We aim to minimise time wastage and make the user journey efficient. From an ethical standpoint, we've kept each user's experience concise and focused. Importantly, no financial loss occurred throughout the entire campaign. We believe that many hardworking individuals have benefited financially from this experience.
What challenges did you encounter during the conceptualisation of this campaign?
The technological aspect posed a challenge because we set a very high benchmark for ourselves. We aimed to ensure that our ads convincingly resembled neuropathy, yet with subtle clues for observant viewers.
Specifically, we intentionally left certain lines on Nora Fatehi’s face visible and misspelt some creatives for the identification of mistakes in the ads. The discount numbers themselves were designed to raise a red flag. Across all our ads and posts on the official page, we strategically placed these ‘easter eggs’ for people to spot and engage with.
The results have been incredibly positive for us, providing valuable insights. For example, we've adopted a strategy of sliding into people's direct messages (DMs), much like how fashion brands announce exclusive offers. Returning to the initial point, we've successfully executed the creative strategy of scammers and the media strategy employed in scams.
We're actively sending ads on WhatsApp, featuring on Google searches, appearing on Meta, making phone calls, and saturating various touchpoints where scammers typically operate. This approach was not without its challenges, but it has proven to be rewarding and a significant learning experience.
Undertaking something as risky as this required a considerable amount of support and faith in our ability to adhere to the rules. A project like this entails potential risks, but with the right intentions and the right team, we are confident in our approach.
How do you anticipate this deepfake approach impacting audience engagement and awareness for the brand?
Already, the signs indicate that people are responding positively. Numerous individuals are actively sharing the campaign, expressing disbelief and questioning how this could happen. Many are sharing our ads to alert others. Instead of saying, ‘I fell victim to this scam,’ they're forwarding messages like, ‘look at this brand it has 80% off.’
Consumers are engaging in the same journey as the brand, realising the potential vulnerability and exclaiming, ‘Oh, I could easily fall for this.’
For us, this is the ultimate litmus test, gauging whether people are embracing the campaign well. All the reviews have been exceedingly positive, with everyone appreciating our efforts.
Even from within the industry, particularly the banking sector, which traditionally tends to be conservative and risk-averse, the response has been favourable. The BFSI sector typically plays it safe, but as we delved into this campaign, there were initial concerns. However, we are delighted that we took this bold step, and the feedback has been amazing.