Digital arrest hardest fraud threat, Google can't fix it alone: Kotsovinos

Digital arrest scams are the toughest fraud challenge today, calling for ecosystem-wide action as deepfakes and AI-driven impersonation make tracking and prevention more complex, says Kotsovinos

New-age frauds and scams, such as digital arrests, are the most complex part of the puzzle to crack in the entire scam and fraud ecosystem. Solving this issue will require coordinated efforts from all stakeholders, including companies other than Google, Evan Kotsovinos, the company’s vice president of privacy, safety and security, tells Aashish Aryan in an interview. Edited excerpts: 

The aspects of safety and vulnerability are very different in markets like India compared to more digitally mature markets such as the EU, the US, or the UK. What are the safety measures that you adopt in India? 

There are a couple of examples. We are working to catch and prevent scams happening through short message service (SMS) or voice calls. Another method is safe browsing in Chrome, which, even if you click on a malicious link, does its best to prevent you from visiting the site. 

There are large-scale attempts to manipulate paid search results for fraud. We invest heavily in preventing that from happening. Similarly, in advertisements, we use artificial intelligence (AI) to avoid fraud of all kinds. 

Digital arrests are sort of the cutting edge of the problem, and to my mind, the most challenging part of the ecosystem right now. It is a very complex problem, and Google cannot solve it alone. 

Why are digital arrests so hard to track or solve, even with all the advancements in technology? 

Many of these digital arrest calls originate from deep fakes or impersonations of law enforcement officials made through apps that are not based on our AI models. There is no single technology right now that solves this problem. 

This is an ecosystem problem, and we have to ensure everyone embeds information in AI-generated media. There are some early technologies that attempt to detect AI-generated content, even if it is not watermarked. Those are in early stages, but we are investing in research in that area. 

Scams and scammers are also likely to become more sophisticated and faster with the advent of AI. What can companies do to stay ahead? 

In this arms race against scammers, we need to keep improving. At the same time, AI can be used to drive greater safety. We are investing in software that can detect vulnerabilities in other software, devise a patch for that, and fix the issue all on its own. We have developed an AI agent that finds these vulnerabilities. Since we launched it last November, the agent has found more than 100 previously unknown vulnerabilities. I am optimistic that ultimately, AI will make the world safer. 

Speaking of privacy in India, do you think the users’ rights will improve now that the Digital Personal Data Protection (DPDP) Act has been operationalised? 

I do hope that the public here becomes more engaged in the debate on privacy and stands up for what is the right outcome for them, and their data.