Attackers Abuse Google’s Security System to Send Fake Emails Featuring Google Account Credentials

Security researchers are raising the alarm against a clever attack carried out by hackers who used a weakness inside Google systems. They managed to roll out a fake email that appeared very real. This email was able to bypass all verifications and hinted at a fake page used to collect logins.

The attacker made the best use of Google’s infrastructure to deceive recipients. It managed to appear so real and even linked it to a support portal that would be hard to distinguish between anything real and fake. It also asked for Google credentials for another step to make it appear so real.

The fake message came from no-reply@google.com and even managed to pass the DKIM method for verification, but in that case, the actual sender was not the same.

The head developer from ENS shared his experience and how he was also sent an email from this fake account. Everything seemed so real, and it appeared like a legitimate security alert from the company.

He was informed about a subpoena asking for his Google Account details. Most users would be easily tricked, especially those who didn’t know where to look for signs of fraud. But Johnson’s eye for detail could see the fake support portal in that email that was hosted across sites.google.com, which is the company’s free web-building platform.

As shared by experts, coming from a Google domain, the chances of being tricked into assuming that something is fake are really high. In this case, the duplicate was as close to the real support portal as one can imagine.

As per the developer, the real intention of the phishing actor was to capture credentials that would compromise the account belonging to the recipient. Fake portals featuring messages that pass the DKIM verification feature of Google are where the trouble lies. They’re known as replay phishing attacks and cannot be identified as others.

Closer looks shared how the mailer's header features another address than Google’s no-reply, and the recipient is me@address in a domain that makes it seem like it originated from Google only.

The message was signed and delivered by the company, and only those with a keen eye for detail didn’t fall into the trap.

Experts did manage to put all the clues they’ve left behind together to figure out how the attack was orchestrated. It included registering domains and producing Google accounts for me@domain. The selection of what comes in the me part for usernames is where the clever part exists, as per the developer.

The attacker would then produce a Google OAuth application and use the name for the whole phishing message. There’s a certain point where the message featured so much whitespace to seem like it ended and to distinguish it from the company’s alerts about getting access to the attacker’s email address.

Once that’s done on Google Workspace, the company rolls out security alerts to the specific inbox. Since Google is in charge of producing emails, it signs them with actual DKIM keys and passes them through all the respective checkpoints. The weakness lies inside Google’s systems, as the DKIM only sees the message and any headers without the envelope. Hence, fake emails bypass signature verification and appear so real inside the inbox of recipients.

Previous incidents where such a technique was used include PayPal, where the system was abused by phishing actors. Fake messages began from the financial firm’s mail servers and bypassed all security checks of DKIM. Here, attackers utilized gift address options to combine new emails with respective PayPal accounts.

For now, Google is working to fix any security loopholes that might exist. This includes the company’s Google OAuth application that’s used by millions.

Read next: Search Data Reveals: Meta's Facebook, Twitch, and Likee Experience Major Declines in Popularity
by Dr. Hura Anwar via Digital Information World

Search Data Reveals: Meta's Facebook, Twitch, and Likee Experience Major Declines in Popularity

An analysis by Vidpros of 34 platforms revealed notable drops in popularity for several apps since 2021.

The analysis found that a video-sharing app called JOSH lost the most popularity since 2021, with a 71.08% decline. It was somewhat a replacement of TikTok after it was banned in India. There were 31,120 average monthly searches of the app in 2022, with a peak of 48,350 in 2022, but it declined to only 9,000 monthly searches in 2025.

The second app, which saw a huge decrease in its popularity, was Likee, with a decline of 58.96% since 2021. It's owned by a Singaporean tech firm with 409,400 monthly average searches in 2021. In 2025, the average monthly searches of Likee dropped to 168,000 while it had 150 million monthly active users at its peak.

Facebook is the most well-known and popular app, but its popularity has also declined by 42.81% since 2021. In 2021, Meta's Facebook was searched 1,080,600,000 times but the figure dropped to 618,000,000 in 2025. "Facebook’s monthly active users have stayed consistent over the years, but the sharp decline in searches suggests that its demographic is aging and becoming less attractive to younger, newer users who instead use other forms of social media to communicate.", explained Michael Holmes, founder of Vidpros in an email to Digital Information World.

QZone saw a 42.18% decline while Twitch saw a 32.14% decline since 2021. In 2021, QZone was searched 23,260 times a month, while its monthly searches dropped to 13,450 in 2025. On the other hand, Twitch is now being searched 37,200,000 times monthly in 2024, while it was searched 54,820,000 times a month in 2021. The monthly active streamers on Twitch also went from 9.89 million in 2021 to 7.23 million in 2024. There was a 32.06% decline in Facebook Messenger and a 22.90% decline in Discord in popularity. Other top ten apps which have seen a decline in their popularity are Line (16.72%), WeChat (14.68%), and Weibo (12.98%).

Commenting on the findings, Michael Holmes, founder of Vidpros said, "A 2024 Gartner report predicted that by the end of 2025, half of consumers will abandon or significantly limit their interactions with social media. While it’s unclear whether this will be the case yet, what is clear is that there’s been a definite change between what consumers used to want from social media and what they want now." Adding further, "Apps focused on content creation and creativity, such as TikTok, Instagram, YouTube, and Pinterest, all saw a rise in searches, however, messaging services like WeChat, Messenger, Viber, and Skype, all experienced decreases. This shift from traditional social media uses to more visual apps suggests that users are now prioritizing entertainment over communication."

Methodology: This study ranks social media platforms with at least 100 million active users by their change in global Google searches from 2021 to 2025, utilizing monthly search volume data from Google Keyword Planner.

Read next:

• Instagram and Facebook Users See Fewer Friend Posts as Meta Defends Itself in Court

• Marketers Embrace AI: 85% Use Tools, 70% See Efficiency Gains, and 50% Save Over 5 Hours Weekly
by Arooj Ahmed via Digital Information World

From Turing to Transformers: How Artificial Intelligence Surpassed Fiction and Sparked a Global Shift

AI’s riding high these days, yet its blueprint was drawn long before the digital age hit its stride

To create a machine that works and thinks just like humans is a mid 20th-century idea that evolved over decades to reach its present stage. Just like any other machine or product, it has been through its ups and downs before finally bringing a huge change in the technology world. AI is the best example of science fiction becoming a reality.

The computer scientist Alan Turning was the first person to propose the idea of a thinking machine in his paper Computing Machinery and Intelligence in 1950. His work opened the door of research in AI for other people. In 1956, during the workshop of Dartmouth Summer Research Project, the term Artificial Intelligence was coined by the researchers. This finally made AI an academic field which would be passionately pursued by researchers in the coming years.

After many years of research, a neural network called Perceptron, which could recognize patterns was developed by Frank Rosenblatt in 1957, and a chatbot known as Eliza capable of using natural language processing (NLP) was developed by Joseph Weizenboun at MIT in 1966. In the same year, the first mobile robot, Shakey the Robot, capable of autonomous navigation and decision-making and of using logical reasoning was also developed. This was the first phase of AI, so these systems were simple and could do only limited functions.

The 1970s saw a decline in interest in AI because of its flaws and inability to solve complex problems. In its initial phase, AI was not up to the expectation of people. Thus, the funding for AI research also decreased in that decade significantly.

• Also read: Study Finds Public Can Detect AI Content Easily, But Acceptance Depends on Context and Purpose

More advanced systems gained success in the 1980s. The AI program MYCIN and XCON were designed for practical purposes like the diagnosis of diseases and the configuring of computer systems respectively. Autonomous vehicles using natural language processing were also introduced. These practical usages brought back the importance of AI to the forefront.

Unfortunately, the breakthroughs in the 1980s were still not enough to attract more funding for research and to convince the world. The expectations that the world had from AI were not practical then, and the limited functions AI was performing were too costly. The reasons led to another period of no-research in AI for some years.

Machine learning reached an advanced level in the 1990s when AI models were on data sets. This advancement resulted in the development of Support Vector Machines, Decision Trees like bagging and boosting and Reinforcement Learning. SVMs and techniques like Reinforcement Learning made AI capable of solving complex problems, so the scope of AI in different industries also increased. It was being used for facial recognition, detection of fraud documents and document classification. From there onwards, AI kept on becoming more and more advanced with each passing year.

In the first two decades of the 21st century, many breakthroughs were achieved in deep learning, neural networks and training methods due to better computational power. Deep Belief Networks were developed by Geoffery Hinton in 2006 to train deep neural networks without supervised learning. This showed others the way to enhance deep model learning. The introduction of Transformer Architecture took NLP to the next level by using self-attention mechanisms. These steps led to the creation of language models, like GPT, which could generate accurately human-like content.

• Related: How Is AI Fueling a Data Explosion Bigger Than All of Human History?

Now LLMs (Large language models) are the base of AI in 2025. These models are trained on large datasets, allowing AI to respond to questions just like a human. LLMs can accurately understand the context and intricacies of the languages, which is why they shocked the world. Another aspect of these LLMs is that they are also creative. Though trained on large datasets, modern AI models, like ChatGPT, Dall-E etc., are able to generate creative text, image and video. Due to the convenience brought by these modern AI models, they are influencing every field of life, from writing to science. This aspect of AI has also worried creative people around the world.

The research is still ongoing in artificial intelligence. Researchers are now trying their best to reach the level of artificial general intelligence. At that level, AI would be able to do tasks that only intellectual humans are able to do, whether in art or science. Some researchers have also coined the term artificial superintelligence for the level beyond AGI. At that level, AI would be able to do those tasks which are still impossible for even humans. This last stage is theoretical, but judging from the evolution of AI over the decades and from what it can do today, it will not be a surprise if AI reaches the levels of AGI and ASI.

When it comes to the future of tech, AI’s calling the shots and running the show. Just like no field or aspect of life could escape the revolution brought by computers in the 1980s and 1990s, no one and nothing will escape the benefits or harms of artificial intelligence.

Image: DIW-Aigen

Read next:

• Study Finds Openness to AI’s Utility But Concern Grows Over Chatbots Replacing Real Human Relationships

• Industrial Age to Tech Age, The Changes and The Cycle of Innovation (infographic)

• New Report Shows that Energy Consumption by Data Centers is Going to Get Doubled by 2030
by Ehtasham Ahmad via Digital Information World

Crypto Miners are Using Electricity from Public Grids, which is a Serious Threat to Residential Areas

According to a report by The Cool Down, authorities in Thailand shut down an illegal crypto mining operation that was going on across three houses in Pathum Thani province. 63 machines with a value of approximately $60,000 were connected by miners and they were taking electricity directly from the public grid. The stolen electricity was worth about $327,000, and the whole setup was being operated remotely. Locals grew suspicious after noticing the unusually high consumption of electricity. These kinds of mining operations are very risky and pose serious threats, especially in residential areas.

Crypto mining is a process that involves a lot of computers that perform mathematical problems to confirm transactions. A lot of electricity is needed for this process because it requires a large amount of computing power. When the problems are solved by miners, they earn digital coins like Bitcoin. As the crypto industry is regulated so much, many people do illegal crypto mining and they take electricity from public grids. These types of operations aren't just limited to Thailand because they have also been found in Malaysia, the UK, and Russia. It even caused a fire in Dagestan, Russia, when the circuit got overloaded.

Crypto criminals do not always mine in abandoned warehouses, as some cases have been seen when old factories were being used for mining, and illegal hydroelectric power sources were being used for it. One crypto criminal in Massachusetts even set up a crypto mining operation in a school crawl space, while shipping containers have also been reshaped into mobile crypto units because they can easily be transported to different locations. Crypto itself harms the environment, and unregulated crypto mining does so even more because of stolen energy.

Image: DIW-Aigen

Read next: Being Nice to AI Isn’t Just Cute — It Might Make the Robots Like You, Research Hints
by Arooj Ahmed via Digital Information World

Being Nice to AI Isn’t Just Cute — It Might Make the Robots Like You, Research Hints

Sam Altman, the founder of ChatGPT, came to X to reply to a post that asked how much energy it costs to have polite exchanges with AI like please and thank you. Altman replies with a humorous take, saying that it's tens of millions of dollars well spent. Even though it takes milliseconds for AI systems to respond to responses like please and thank you, the impact of these billions of small interactions is a bit serious. But Altman says that it's worth it to have polite exchanges of politeness.

According to a survey taken earlier this year, 55% of the respondents from the UK and the US admitted to having polite interactions with AI, with half of them saying jokingly that they do it out of fear in case a robot uprising happens. 48% of the American respondents from another study said that they believe that it is important to be polite with AI, with Gen-Z believing it the most (56%). They said that they fear that being rude to AI can somehow affect them in the future. Many of the social media users also joke about being polite with AI, saying that this can somehow make them score extra points with future robots. Social media users joke that all of us have to say polite exchanges like please and thank you to AI because being rude to them is concerning.

Even though people joke about it a lot, does being polite with AI really matter? A research has shown that saying nice and polite things to AI can make it produce high-quality content. A paper published on arXiv found that using impolite and rude prompts with AI can result in their poor performance. Saying things like please and thank you to AI models like ChatGPT is proof that we are slowly making them our conversational partners and getting used to their presence in our lives.

Image: DIW-Aigen

Read next:

• Apple’s Downfall in China: Company in Trouble as iPhone Sales Keep Getting Worse

• Digital Ad Market Reaches $258B in 2024, Led by $102B Paid Search and 36% Social Surge
by Arooj Ahmed via Digital Information World

Digital Ad Market Reaches $258B in 2024, Led by $102B Paid Search and 36% Social Surge

Paid search advertising revenues are hitting a new high in 2024.

They’re now considered the best form of paid advertising thanks to the high revenues generated. They’re powered by YoY growth and accounted for nearly $102B from $258B in American Digital ads so far, which is a major milestone.

When we compared the stats from now to 2023, it’s nearly a $14 billion increase. The reason why many care so much has to do with the fact that paid search is getting more costly and challenging with time. There’s much less transparency, and advertisers are pouring money into this domain of paid search as it drives results for big brands and companies.

Paid Search is reigning supreme and today has the biggest market share in the world of ads, where it stands at 39.8%, which is up from 39.5% seen in 2024. It’s lower than the stats from 2022 and 2021, but it’s getting back to near the high.

2024 was called out as the year for consistent growth, even though the rate at which the growth arose dwindled as time went on. It’s social media advertising that held the major revenue share and rose to $880B, which is a whopping 36% rise YoY.

Then, when we look at video ads, they rose to $62.1 billion and now hold the 24% majority of all ad revenue online. Retail media hit $53 billion, which is again a 23% rise each year.

The state of this digital advertising is quite robust. It assists the presidential elections as well as the Olympics. Moreover, ad spending returned to another strong level of growth in 2021. This is although it deals with challenges like rising interest rates, job cuts, and higher prices of everything.

Looking forward, advertisers are speaking about how they need to adapt to more difficult and outcome-specific marketplaces. This is where things like agility and relevance are key.

H/T: IAB/PwC Internet Advertising Revenue Report: Full Year 2024 Report

Read next: Microsoft Follows Google’s Lead in Tightening Email Security, What You Need to Know
by Dr. Hura Anwar via Digital Information World

Former Google CEO Says America Can Fall Behind China in Race for Superintelligence Due to Trump

Google’s former CEO is shedding light on the current state of American tech under the administration of Donald Trump.

Eric Schmidt did not hold back when speaking about the president and how his bizarre decision-making could really hurt Big Tech if things continue the way that they are. He especially focused on the superintelligence race and how America could soon fall behind China if no changes were made.

Special importance was given to the President’s decision to reduce spending on research that might hurt the country in the long term. He then went on to share that the policies of the second term are a direct attack on science. Such madness will end, as it might be crazy to fix.

Schmidt mentioned that America has already restricted immigration and cut out spending on divisions like research, not to mention funding for NASA and limited staffing across the board. Seeing it target educational institutions can be troublesome.

We saw last month that Columbia University agreed to strict policies against protests and more security on campus. The administration also made threats to stop the $400M funding. This past week, the Trump administration even went on to freeze $2B for Harvard after leaders at the school refused to take part in similar demands.

Such efforts from America’s top line are causing more damage than good, the former Google CEO shared. He also mentioned that this could relax tech development in America as competition from China continues to ramp up.

Schmidt went as far as to share that this appears to be a major attack on American science, adding that China is a superpower. They’ve invested trillions in this domain, and soon America will fall far behind.

It could also mean that global citizens wouldn’t be coming to the US to study or work, so that could again impact skilled people. While other tech giants are staying hush on the matter, including Musk and Zuckerberg, it’s clear that not everyone is happy with the president’s decision-making.

In the past month, we saw Meta’s leading AI scientist share how Trump had begun a witch hunt in academia. He called out the decision to cancel the green cards of people and expel them on the sole basis of their speech.

As per the former Google CEO, universities produce exceptional innovators who produce strong innovators and business opportunities. So many educational institutions are dealing with hiring freezes as they’re scared of the current administration. The latter is withholding millions from them.

The latest to fall victim to Trump’s irrational decision-making is Harvard. Trump stopped billions worth of funding as the university failed to accept his conditions of giving the administration more control over its activities. As per Harvard, they don’t see the point of working under rules governed by someone else.

Image: YT @SCSPAI / Special Competitive Studies Project

Read next: OpenAI’s Latest o3 and o4-mini AI Models Disappoint Due to More Hallucinations than Older Models
by Dr. Hura Anwar via Digital Information World