Latest news with #GPTBot
Hans India
01-08-2025
- Business
- Hans India
Billions lost to brands as AI-powered ad fraud exploits every click
The Economic Times says the illicit supply chain is 'fuelled by advanced bots, airtight automation, and low-cost black-market engagement'. Industry professionals in adtech are seeing 'a 86% YoY surge in invalid bot traffic by late 2024, with sources tracing this activity to automated crawlers and scrapers such as GPTBot and ClaudeBot skewing campaign performance and eroding trust throughout the digital marketing fraud ecosystem,' a new release said. Big names have suffered collateral damage, with one major OTT platform said to have seen 100% bot traffic and a 52% click through rate (CTR), clear signals of inauthentic activity. 'It has become very easy for fraudsters to customize a programmatic ad fraud campaign to pump up high-level engagement metrics such as reach and conversions,' Shailesh Dhuri, CEO, Decimal Point Analytics, told the publication. 'In fact, you will really only see a campaign's effectiveness once you filter out numbers from these bots.' Cyber criminals are simulating organic behaviour, such as scroll patterns, dwell times on elements, jitter and click timestamp patterns and device fingerprint characteristics, to avoid 'traditional bot and fraud filtering mechanisms.' Others are finding that some analytics dashboards now include 'fraud-adjusted views to surface ads viewed by bots and provide clear insights into real ROI'. AI-led digital fraud is a common weapon in the ad fraud arsenal, but click-jacking and AI is now being weaponized to create deepfake ads, site cloners, fake 'news,' and, especially popular among scammers, fake influencer content. Celebrity videos created by Generative AI—which purport to be endorsements, movie promotions or other forms of marketing communications—are already pasting together scam campaigns that are beating detection tools and duping consumers into engaging with sponsored content. In China, Egypt, Bangladesh, Kenya, and other markets, 'brand loss due to ad fraud' are booming. A thousand reel views on Instagram, for example, can cost an ad fraud 2025 just about a dollar, The Economic Times points out.
Time of India
01-08-2025
- Business
- Time of India
Click & Bait: AI-led fake ads skim Billions from Brands
New Delhi: Want to boost your Instagram reel views? You can now buy 1,000 views for a dollar. From China to Egypt the booming black market of fake clicks or 'click farms' are gaming the system to artificially inflate website traffic metrics and mislead advertisers. Similar ad fraud tactics like traffic exchange credits and proxy traffic services are burning advertising dollars. For instance, in a traffic exchange networks, users agree to view each other's websites in order to increase page views or ad impressions, but they do not lead to any real engagement or conversions. And AI is only making such scams more sophisticated with tricks like ad hijacking, cloaked sites and smart redirects. "We're seeing an avalanche of AI-led fake traffic," said Dhiraj Gupta, founder of digital fraud detection firm mFilterIt. "Tools like ChatGPT can now generate thousands of sites in hours, and platforms like offer millions of bot hits for a few thousand dollars." Nearly 22 per cent ($84 billion) of online ad spend is lost to ad fraud in 2023 and is expected to reach over $170 billion by 2028, according to Juniper Research. According to research by cybersecurity firm Imperva, bots now account for over 51 per cent of all internet traffic, surpassing human activity for the first time. Nearly 37 per cent of all internet traffic was characterised by bad bots for activities like account hijacking, data scraping and mimicking human behaviour to avoid detection. Ad measurement company DoubleVerify (DV) found an 86 per cent year-over-year increase in invalid bot traffic in late 2024, which they attribute to the rise in AI-powered crawlers and scrapers such as GPTBot, ClaudeBot. While these bots lack malicious intent, they can distort campaign metrics, inflate impression counts, and undermine trust in the digital advertising ecosystem, the report said. Advertisers are largely unaware of the scale of the problem, Gupta emphasised. Even top-tier brands have been caught off-guard by such bot attacks. For instance, a major OTT platform crashed due to 100 per cent bot traffic and 52 per cent click-through-rate, a clear sign of made-up engagement, he recalled. During the India-Pakistan tensions, several YouTube videos were injected with hate speech ads maligning the platform's and brand's reputation, he said. At a base level, the ad-fraud economy is booming because many brands still track success by the number of clicks and impressions registered, which are easy to manipulate. "Fraudsters optimise for surface metrics," said Shailesh Dhuri, CEO, Decimal Point Analytics. "True effectiveness only shows when you remove the inflated numbers from bots."
Time of India
01-08-2025
- Business
- Time of India
AI-led ad frauds skim billions from brands one click at a time
ETtech Want to boost your Instagram reel views? You can now buy 1,000 views for a dollar. From China to Egypt, the booming black market of fake clicks or 'click farms' are gaming the system to artificially inflate website traffic metrics and mislead ad fraud tactics like traffic exchange credits, proxy traffic services are burning advertising dollars. For instance, in a traffic exchange networks, users agree to view each other's websites in order to increase page views or ad impressions, but they do not lead to any real engagement or conversions. And AI is only making such scams more sophisticated with tricks like ad hijacking, cloaked sites, and smart redirects.'We're seeing an avalanche of AI-led fake traffic,' said Dhiraj Gupta, founder of digital fraud detection firm mFilterIt. 'Tools like ChatGPT can now generate thousands of sites in hours, and platforms like offer millions of bot hits for a few thousand dollars.'Nearly 22% ($84 billion) of online ad spend is lost to ad fraud in 2023 and is expected to reach over $170 billion by 2028, according to Juniper to research by cybersecurity firm Imperva, bots now account for over 51% of all internet traffic, surpassing human activity for the first time. Nearly 37% of all internet traffic was characterised by bad bots for activities like account hijacking, data scraping and mimicking human behaviour to avoid measurement company DoubleVerify (DV) found an 86% year-over-year increase in invalid bot traffic in late 2024, which they attribute to the rise in AI-powered crawlers and scrapers such as GPTBot, ClaudeBot. While these bots lack malicious intent, they can distort campaign metrics, inflate impression counts, and undermine trust in the digital advertising ecosystem, the report are largely unaware of the scale of the problem, Gupta emphasised. Even top-tier brands have been caught off-guard by such bot attacks. For instance, a major OTT platform crashed due to 100% bot traffic and 52% click-through-rate, a clear sign of made-up engagement, he recalled. During the India-Pakistan tensions, several YouTube videos were injected with hate speech ads maligning the platform's and brand's reputation, he a base level, the ad-fraud economy is booming because many brands still track success by the number of clicks and impressions registered, which are easy to manipulate. 'Fraudsters optimise for surface metrics,' said Shailesh Dhuri, CEO, Decimal Point Analytics. 'True effectiveness only shows when you remove the inflated numbers from bots.'The company has observed a spike in bot sophistication enabled by simulated mouse movements, and even fake conversion paths. The AI models analyse everything from scroll behaviour to entropy in click timing and device fingerprint inconsistencies, Dhuri explained. In fact, nowadays, fraud-adjusted dashboards filter out bot traffic to provide more accurate ROI the threat goes beyond bots clicking ads. Scammers are increasingly using generative AI to craft deepfake ads, clone websites, and create mechanical influencer videos.'There have been fake betting ads using AI-generated videos of Sachin Tendulkar, Shah Rukh Khan, and Virat Kohli which look alarmingly real, even fooling detection systems,' said Pratim Mukherjee, senior director of engineering, are increasingly employing dynamic short-form content, such as reels and shorts, enhanced withAI-generated celebrity voices and movements, to create highly convincing attacks that deceive social media users, Mukherjee cyber protection firmware company detected over 36,000 fake Amazon sites and more than 75,000 impersonation scam texts around Prime Day sale in July this also highlighted the need for a native traffic verification framework, something that is tailored to the country's massive scale and ecosystem.'You need AI to fight AI,' said Gupta. 'Our models track traffic patterns as a unit, looking for clusters that behave identically and deviate from normal human activity. But too often, brands don't even know they're being hit.'
Indian Express
02-07-2025
- Business
- Indian Express
India must act: Publishers sound alarm on AI's ‘theft' of news content
As leading publishers in the US and UK clamp down on artificial intelligence (AI) companies scraping news content without consent, Indian digital media houses are ramping up pressure on the government to step in and protect journalistic work from being 'exploited' by commercial AI models. The development follows a sweeping move by major global players, including the Associated Press, The Atlantic, Sky News, Time, Buzzfeed, Conde Nast, and DMGT, to block AI bots from crawling their websites by default. The effort is being supported by Cloudflare, one of the world's largest internet infrastructure firms, which has announced a new system that gives publishers granular control over AI access to their sites. Indian publishers say the problem of unauthorised AI scraping has reached alarming proportions with no legal safeguards, licensing systems, or enforcement mechanisms in place. 'The situation in India is becoming increasingly untenable,' said a spokesperson for the Digital News Publishers Association (DNPA), a key industry body. 'While global players are waking up to the importance of permission and fair compensation, Indian news content continues to be freely exploited without dialogue or safeguards. We urge the Government of India to take immediate steps to take necessary measures against such unauthorised and rampant data scrapping,' the spokersperson said in a press statement. According to Cloudflare, OpenAI's GPTBot alone accounted for nearly 30 per cent of all AI-related web scraping in May 2025, a sixfold increase from just a year earlier. Other major scrapers include Meta's External Agent and Anthropic's ClaudeBot. Cloudflare's new approach gives publishers the option to allow or block AI crawlers and even tag them based on intent, whether the bot is scraping for model training, search indexing, or inference purposes. Crucially, the company is also testing a 'pay-per-crawl' model that could allow news outlets to charge AI firms directly for accessing their content. 'This is a game-changer for publishers,' said Roger Lynch, CEO of Conde Nast. 'When AI companies can no longer take anything they want for free, it opens the door to sustainable innovation built on permission and partnership.' But in India, publishers have few tools at their disposal. Unlike in the US and UK, where AI companies now face increasing legal scrutiny and regulatory pushback, Indian media houses are still operating in a grey zone. The DNPA and other digital publishers have now jointly called on the Ministry of Electronics and IT (MeitY) and the Ministry of Information and Broadcasting (MIB) to urgently act on the following demands: 'India has the potential to be a global AI leader – but not by trampling on the rights of its own creators,' said a senior editor from a leading digital news platform. 'We must innovate responsibly, with laws that value original content and protect public trust.'
