AI: A Double-Edged Sword in Online Advertising

The rise of generative AI tools has intensified the ongoing battle against online spam and scams. 
 
As scammers exploit this technology to create deceptive content, tech giants like Google are also leveraging AI to combat these threats.
  
AI-generated scams have grown rapidly, evolving from fake herbal remedy advertisements to sophisticated videos that mimic the voices of well-known personalities. 
 
While the problem of deceptive online content is not new, advancements in artificial intelligence have dramatically amplified its scale and speed. 
 
Today, both malicious actors and others can create convincing fraudulent material much more quickly and in far greater volumes than before.
 
According to the FBI's latest Internet Crime Report, over 22,000 AI-related scam complaints were filed last year, resulting in losses exceeding $893 million. 

 
In response, Google has enhanced its AI systems, particularly its Gemini technology, which successfully identified over 99% of policy-violating ads before they reached consumers.
 
In 2025, Google blocked more than 8.3 billion ads, including 602 million associated with scams. 
 
The company suspended around 24.9 million advertiser accounts, over 4 million for scam-related activities. 
 
Despite this, Google remains a leader in digital advertising, generating over $200 billion in revenue last year.
 
Keerat Sharma, Google’s vice president of ads privacy and safety, emphasizes the effectiveness of AI in analyzing vast amounts of data to assess advertiser intent. 
 
This capability has significantly reduced incorrect suspensions by 80%, ensuring legitimate businesses can continue advertising.
 
Google's multi-faceted approach includes rapid analysis of ad content, enabling real-time defense against spam and scams. 
 
However, the landscape remains complex, with various forms of deceptive ads emerging at a faster rate.
 
Experts warn that the conflict between AI-generated scams and AI defenses will persist. 
 
As technology evolves, the struggle will increasingly resemble “AI versus AI,” highlighting the challenges of managing this fast-growing issue effectively.