How GenAI is stirring chaos in the digital world

New Delhi: GenAI is turning the digital world upside down with its conversational ability. Marketers love it for crunching data and crafting personalised content at lightning speed; consumers rely on it for quick info fixes; and publishers churn out content in a flash.

But behind that friendly AI chat lie hidden dangers — marketers may lose genuine connections with consumers, users could unknowingly spill personal data, and publishers might miss out on cashing in. It’s a digital game-changer, but one with a few tricks up its virtual sleeve!

The reliance on generative AI introduces concerns about bias and accuracy in marketing. 

"Gen AI models, which learn from historical data, can perpetuate existing biases. This is a significant issue in marketing, where innovation is crucial. AI’s reliance on historical trends can limit creativity and lead to confirmation bias," highlighted Roy Menezes, partner and chief creative officer, Centrick.

"A lot of Gen AI models rely on historical data, which can lead to confirmation bias. This limits the potential for new and innovative marketing strategies and can stifle creativity. This is particularly relevant in marketing, where fresh, creative approaches are essential," Menezes further explained.

Alekhya Chakrabarty, Vice-President of Marketing and Growth, Unstop, seconded Menezes and said, "Generative AI models rely on existing data inputs and can perpetuate confirmation bias. They often fall short in creating truly innovative content because they are constrained by historical data trends.” 

"Today, as we've come out of the pandemic, the last five years of data is completely insignificant again. So how is a bot making an informed decision? The minute an unfamiliarity sweeps in, I don't think a bot can function," Menezes observed, emphasising the limitations of Gen AI in handling dynamic and unpredictable market conditions.

"You need to understand that when content is being generated, it is only between me and the machine, right? It's called reinforcement learning feedback. When GPT responds, there is an option of thumbs up or thumbs down. So whatever feedback you are giving, it becomes part of how it may respond in the future. Since all these systems are black boxes, we do not know how much impact it will have on the performance of GPT by using this information," commented Garima Saxena, Senior Research Associate at The Dialogue.

Therefore, Chakrabarty emphasised the need for human oversight to complement AI's capabilities, suggesting that Gen AI should be used for direction, with human intelligence applied to generate novel ideas.

"The simple thumb rule is to use Gen AI for direction and then apply human intelligence to generate new content," Chakrabarty advised. This approach ensures that AI serves as a tool for inspiration rather than a sole creator, preserving the creative aspects essential in marketing.

“The usefulness of AI greatly depends on its application. We need a nuanced understanding of Gen AI—not just about where to use it, but more about where and how not to use it—sensitisation around Gen AI’s role is essential," advised Chakrabarty.

For smaller businesses, over-reliance on AI tools poses challenges. "Small-scale businesses may depend too heavily on AI without the resources to verify sources or ensure cultural sensitivity," Garima Saxena, Senior Research Associate at The Dialogue, noted. "This can lead to issues with authenticity and relevance, affecting their ability to connect with consumers effectively."

According to Dhruv Garg, tech and policy lawyer, it’s conceivable that amid the prompts and answers generated by AI tools like ChatGPT, Bard, and Copilot, advertisements could soon be integrated directly into these interactions.

Throwing light on the potential impact of advertising on conversational AI platforms, Gard said, "This shift would concentrate more power within tech platforms hosting conversational AI, potentially diminishing the bargaining power of publishers. Even advertisers might find themselves in a weaker position, having to negotiate with the dominant tech platforms.” 

Even from an ethical and social perspective, AI conversations can easily produce misinformation. For instance, deep fakes are a big conversation at this point. Apart from that, there’s also a matter of increased competitiveness—big tech always has better data on us. 

“Generative AI penetrating into our workspace or personal life is going to make dependency higher and higher. Giving way to the dead internet theory—how our world is being filled with more and more AI-generated data," said Saxena of The Dialogue added, addressing the complexities introduced by AI technologies.

Google’s attempt to implement its AI overview feature, for instance, encountered significant issues, including inaccuracies and safety concerns, leading to its rollback. "The challenges with AI implementations, such as Google’s AI overview, highlight the difficulties in ensuring accuracy and safety in these new interactions," Garg noted.

Not just marketing; GenAI-led tech platforms are also impacting publishers’ businesses. For example, a significant debate is whether tech giants should compensate publishers for using their content to train AI systems. This intersects with intellectual property rights and data usage regulations. 

Herein, comes India's efforts to implement the Digital Competition Bill. This legislative measure is designed to tackle the growing dominance of big tech firms in the digital landscape, aiming to enhance market competition and regulate the activities of these major players. The bill seeks to curb monopolistic practices, ensure fair competition, and provide greater transparency in data practices. It also addresses consumer protection by aiming to prevent unfair practices and enhance oversight of digital markets.

The regulatory landscape is crucial in shaping the future of digital advertising, particularly with regulations such as the Digital Personal Data Protection (DPDP) Act in India. "The DPDP Act was introduced rapidly, inspired by GDPR, but it needs to go through multiple levels of conversation and adjustment to effectively address the new privacy dynamics," explained Garg. 

This legislative effort is particularly relevant as businesses and advertisers navigate the shift from third-party to first-party cookies and adapt to the integration of conversational AI technologies. The Act underscores the importance of securing user consent and maintaining transparency, which are pivotal in the era of GenAI where data collection and usage are becoming increasingly complex.

This legislation will significantly impact how businesses collect and use data, emphasising the need for ongoing dialogue and adaptation. “As companies adjust to the DPDP Act, they must also navigate the evolving landscape of data privacy and ethical AI use, ensuring that their practices align with both regulatory requirements and consumer expectations,” highlighted Saxena.