Google’s Efforts in Combatting Harmful Advertisements in India
Google has successfully removed almost half a billion harmful advertisements in India during the previous year, highlighting both the extent of online misconduct and the increasing significance of artificial intelligence in managing it.
AI Takes on a Proactive Role
In its 2025 Ads Safety Report, the company reported blocking or removing 483.7 million ads in India for breaching its policies, while also suspending 1.7 million advertiser accounts associated with repeated or significant violations. This enforcement was driven primarily by its AI technologies, notably Gemini.
From Reactive to Proactive Enforcement
Google stated that its enforcement systems are evolving from simply reacting to issues to focusing on early detection, facilitated by AI models that evaluate intentions and behavioural indicators.
According to the company, bad actors are leveraging generative AI to craft misleading ads on a large scale, and Gemini assists in identifying and blocking these in real time. They highlighted that over 99% of non-compliant ads globally are blocked at an early stage.
Global Context and Additional Statistics
Across the world, Google reported removing or blocking more than 8.3 billion ads in 2025.
Major Risks: Financial Scams and Impersonation
In India, most violations fell within a few recurring categories, such as misleading financial services advertisements, infringements on trademarks and copyrights, and the misuse of personalisation tools for targeted deception.
The company highlighted that financial fraud and impersonation remain particularly worrisome, as bad actors continuously adapt their strategies to evade preventive measures, often mimicking reputable brands or exploiting user data for targeted scams.
Action Against Repeat Offenders
Beyond removing individual ads, Google emphasised that suspending advertiser accounts is vital for curtailing repeated violations. In India, Google shut down 1.7 million advertiser accounts for persistently breaching its policies or attempting to bypass its enforcement systems.
Advancements in AI for Scanning Threats
The company’s teams have long employed advanced AI to detect and eliminate fraudsters, with Gemini enhancing these efforts significantly.
The company explained that its models evaluate hundreds of billions of signals, such as account longevity, behavioural indicators, and campaign patterns, to intercept threats before they impact users.
The latest models, in contrast to previous keyword-based systems, have a superior understanding of intent, enabling Google to identify malicious content and proactively block it, even when designed to evade detection.
