AI agents are becoming increasingly popular among hackers to exploit online bank accounts. By 2027, it is estimated that they will reduce the time to take over an account by 50%.

This is how journalist Anton Mous tells it from cybernews.com.

This is the stark and harsh reality that the U.S. research and advisory company Gartner describes in its latest report, Predicts 2025: Navigating Imminent AI Turbulence for Cybersecurity.

AI agents are becoming useful tools for attackers to break online account protections. Therefore, marketers will need to introduce monitoring tools to analyse interactions with AI agents. This also means that cybersecurity companies should accelerate the move towards passwordless, phishing-resistant multi-factor authentication (MFA).

Account takeover remains a persistent attack vector because weak authentication credentials, such as passwords, are collected by a variety of means, such as data breaches, phishing, social engineering and malware. Attackers leverage bots to automate a barrage of login attempts to various services in the hope that the credentials have been reused across multiple platforms.

Technology-enabled social engineering will also pose a significant threat to corporate cybersecurity in the near future, including audio and video deepfake. By 2028, 40% of all social engineering attacks are expected to target both senior executives and the general workforce, Gartner predicts.

Gartner notes that although only a few cases have been reported so far, those that have occurred have resulted in significant economic damage to the affected parties. Therefore, these incidents should be seen as a wake-up call and a signal that companies should increase their efforts to protect their digital environment.

Organisations will need to stay abreast of the market and adapt procedures and workflows to try to better resist attacks leveraging counterfeit reality techniques.

Educating employees about the evolving threat landscape through the use of specific training on social engineering with deepfakes is a key step, as Gartner analyst Manuel Acosta believes.

_____

Aquest apunt en català / Esta entrada en español / Post en français

A recent report by the UK’s National Crime Agency (NCA) has highlighted a growing and alarming threat: sadistic online gangs that exploit and abuse minors. These groups, operating in the anonymity of the web, pose an «unprecedented risk» to teenagers, according to the NCA.

The report highlights that sadistic online gangs have become more sophisticated in their methods of exploitation. They use various social media platforms and messaging applications to contact minors, gain their trust and ultimately abuse them. These gangs may be made up of individuals from different parts of the world, making them difficult to identify and prosecute.

What are these sadistic gangs?

These gangs are made up of individuals who revel in the torture and humiliation of others, especially vulnerable teenagers. They use online platforms, forums and social networks to connect, share abusive material and coordinate attacks.

These gangs are not simply groups of trolls or cyberbullies. They are structured organisations, with defined hierarchies and roles, dedicated to the psychological and, in extreme cases, physical torture of their victims. They use sophisticated techniques to hide their identity and track their prey:

• Anonymity and encryption: They use Tor networks, VPNs and other tools to hide their IP addresses and locations.
• Social engineering: They manipulate their victims through the creation of false identities and the generation of trusting relationships.
• Sharing of abusive material: They exchange videos and images of torture and humiliation, creating a vicious cycle of violence and abuse.
• Coordination of attacks: They plan and execute coordinated attacks, both online and offline, to maximise the damage to their victims.

Their activities include:

• Online sexual abuse: Sharing and production of child sexual abuse material.
• Extreme violence: Promotion of and incitement to acts of physical and psychological violence.
• Public humiliation: Exposure and dissemination of content that is humiliating for victims.
• Extortion and blackmail: Threats to obtain money or compromised material.

The scope of the problem

The NCA report reveals that these gangs are on the rise and that their sophistication and reach are increasing. The COVID-19 pandemic exacerbated the problem, as adolescents spent more time on the Internet, increasing their vulnerability to these predators.

Consequences include:

• Psychological trauma: Anxiety, depression, post-traumatic stress disorder and suicidal thoughts.
• Social isolation: Shame, fear and distrust that hinder social relationships.
• Academic difficulties: Concentration and school performance problems.
• Substance abuse: Attempt to relieve emotional pain through alcohol and drugs.
• Self-harm: As a coping mechanism for emotional pain.

The NCA is working with other law enforcement agencies and international organisations to combat this threat. Strategies are being implemented to improve detection and prosecution of these groups, as well as to educate minors and their families about the risks and how to protect themselves.

Useful resources

• UK National Crime Agency (NCA): https://www.nationalcrimeagency.gov.uk/
• NCA Report: https://www.nationalcrimeagency.gov.uk/threats-2025/nsa-csa-2025
• NCA press release: https://www.nationalcrimeagency.gov.uk/news/sadistic-online-harm-groups-putting-people-at-unprecedented-risk-warns-the-nca
• News article from New York Times: https://www.nytimes.com/2025/03/25/world/europe/online-gangs-teenage-boys-violence.html

_____

Aquest apunt en català / Esta entrada en español / Post en français

Alarm has been sounded in the United States: the use of nitazenes, a class of synthetic opiates, is increasing.[1] This situation should lead to rethinking treatment and response policies for this type of drug. The drug can be found in powder, pills or liquid form and can be mixed with other drugs such as heroin, fentanyl or benzodiazepines.

The response to fentanyl use had already raised problems because the doses of naloxone needed to correct fentanyl overdoses must be much higher than in the case of heroin. Now, nitazenes pose an even greater challenge, because they are more potent opiates than fentanyl. The use of this new drug has already led to several overdose deaths in the United States. The most potent class of nitazenes is forty times stronger than fentanyl and thousands of times stronger than morphine. The danger of this new opioid has meant that its therapeutic use has not been admitted in any case, unlike fentanyl, which has been admitted for clinical use.

There is an unconscious use of nitazenes, as people are unaware that it is present in other drugs they take. There are no tests that detect nitazene levels as there are for fentanyl. Like the latter opioid, nitazenes weaken people who take them. Thus, we can find people who use heroin and have a stable job, meanwhile this situation is much more difficult to find in the case of fentanyl and nitazenes.

Currently, instruments to detect nitazenes are inadequate or non-existent, especially in the United States, where analysis of wastewater is not carried out, as is done in the countries of the European Union. Detection in the blood of people who ingest it is difficult because it disappears very quickly, so detection in wastewater is extremely necessary.

China seems to be the main source of nitazene precursors, which are shipped to Mexico and then illegally entered into the United States.

[1] Vid. ‘Even worse than fentanyls’: nitazenes in the USA – The Lancet

_____

Aquest apunt en català / Esta entrada en español / Post en français

Browser fingerprinting is one of the many tactics used by criminals using phishing to bypass security checks and thus extend the usefulness of malicious attack campaigns.

While legitimate organizations have been using browser fingerprinting to uniquely identify web browsers for the past 15 years, it is now also routinely exploited by cybercriminals: a recent study shows that one in four phishing actors use some form of this technique.

The Director of Operational Intelligence of the Fortra’s PhishLabs, Kevin Cryan, explains in the study that browser fingerprinting uses a variety of client-side checks to establish browser identities, which can then be used to detect bots or other unwanted site visits. In this context, numerous pieces of data can be collected as part of fingerprinting, such as: time zone, language settings, IP address, cookie settings, screen resolution, browser privacy or user-agent string.

Many legitimate providers use browser fingerprinting to detect bots misusing their services and other suspicious activity, but phishing site authors have also realized its advantages and are using the technique to avoid automated systems that may flag their website as phishing. By implementing their own browser fingerprinting controls by loading their site’s content, cybercriminals can hide phishing content in real time.

For example, Fortra has observed that threat actors use browser fingerprinting to bypass Google‘s ad review process. Since Google ‘s review process is semi-automated, the implementation of browser fingerprint checks allowed the criminals to identify when the server was seeing their ad destinations compared to a normal user. If the threat actor suspected Google activity, benign content was displayed. This resulted in phishing reports being rejected by Google because malicious content could not be detected.

The bot fight mode of Cloudflare is an example of a legitimate provider that uses browser fingerprinting techniques to identify and block bots. Whenever a website loads with bot fight mode, the following JavaScript is executed and sends the results to Cloudflare. Depending on the results, a captcha will be presented or blocked.

If the JavaScript is decoded, security teams will see that someone is investigating and can infer from the strings displayed that it is requesting numerous browser properties and testing to see the results.

Once the JavaScript is finished, it generates a fingerprint and sends all the information to the phishing site where the results are analysed by the server. Depending on what it determines, benign or malicious content will be displayed.

This fingerprint contains all browser properties, including information about screen dimensions, operating system, GPU hardware, time zone and many other data points. All this information combined can make it very easy to determine whether the browser is real or an emulator.

In the past, cybercriminals could easily avoid detection by exploiting an intermediate server and changing its UserAgent. However, browser fingerprinting is very effective in identifying these automated systems, allowing authors to modify the content of their site based on the results. Understanding the browser properties being collected by criminals through fingerprinting is critical for security teams to avoid suspicion.

_____

Aquest apunt en català / Esta entrada en español / Post en français

It was around 2009 that U.S. roads began to become deadlier for pedestrians, especially in the dark of night. Fatalities have increased since then and have reversed the effects of decades of safety improvements. And experts still do not agree on the reasons for this situation.

But what is most mystifying of all is that this upward trend in the number of accidents has not occurred in other comparatively wealthy countries. In places like Canada or Australia, a much smaller proportion of pedestrian deaths occur at night, and these deaths, less common in number, generally follow a pattern of decline, not growth.

In an article published in the New York Times authors Emily Badger, Ben Blatt and Josh Katz explain that nighttime pedestrian fatality trends in the U.S. represent a puzzle that neither experts in vehicle design, nor in the study of driver behaviour or road safety and how they interact have been able to figure out.

During 2021, more than 7,300 pedestrians were killed in the United States, three out of four of them during the hours between sunset and sunrise.

This trend adds to what is already a growing gap in highway fatalities between the U.S. and the rest of the world. Speed limits on local roads tend to be higher in North America, laws and cultural prohibitions against dangerous driving may be weaker, and infrastructure in the U.S., in many ways, has been designed to allow speeding cars.

Part of the problem could be that U.S. highways and pedestrians walking along them have been especially susceptible to new potential risks, such as smartphones and ever-increasing vehicles.

Darkness seems to especially threaten pedestrians, or rather, people on foot on U.S. highways. In comparable countries, pedestrians are often more likely to be fatally struck during the daytime.

The most obvious potential risks that have brought about changes in the United States since 2009 within vehicles would be drivers operating a smartphone or a screen, which has become increasingly complex. The curve that marks the timeline of mobile phone development ends up overlapping closely with the increase in pedestrian deaths.

When it comes to other sources of driver impairment, there would be no particular reason to believe that alcohol, speeding or fatigue would necessarily have made a major difference. What has changed is the amount of technology we surround ourselves with behind the wheel.

Smartphones and the way they can distract both drivers and pedestrians are not uniquely American. But there is one aspect that remains distinctly so: the ubiquity in the United States of automatic transmissions that help free up the driver’s hand for other uses. Only 1% of all new passenger vehicles sold in the U.S. in recent years had manual transmissions, while in Europe, for example, they stand at 75%.

Driver behaviour inside the vehicles, of any type, may also have changed during that time for some additional reasons, the researchers suggest. This timeline also overlaps with the rise of opioids and the legalisation of recreational marijuana. But there is still little research on how marijuana affects driving.

_____

Aquest apunt en català / Esta entrada en español / Post en français

Some uses of facial recognition technology raise significant concerns that would merit a swift government response, as expressed by a new report of the National Academies of Sciences, Engineering and Medicine. The paper recommends taking into account federal legislation and an executive order, as well as the attention of the courts, the private sector, civil society associations and other organisations working with facial recognition technology, providing guidance for the responsible development and deployment of the technology.

A powerful and increasingly used tool, facial recognition technology is useful for a wide range of verification and identification applications and offers capabilities to check if someone is who they say they are and identify a person in an image. The systems use trained artificial intelligence models to extract facial features and create a biometric template from an image, and then compare the template features with features from another image or set of images to produce a similarity score. According to the report, the accuracy and speed of these systems have advanced very rapidly in the last decade with the adoption of machine learning.

With few exceptions, the United States currently has no authoritative guidelines, regulations or laws to adequately address issues related to the use of facial recognition technology, the report details. It also states that, even if it does not necessarily violate the rights and obligations included in statutes or constitutional provisions, facial recognition technology may interfere with and substantially affect the values embodied in U.S. privacy, civil liberties and human rights commitments.

According to University of Wisconsin-Madison chancellor Jennifer Mnookin, facial recognition technology creates new and complex legal challenges and raises a variety of different and unresolved legal issues. It also raises complicated social questions about privacy and public and private surveillance, given the very personal implications of the technology.

It is crucial for governments to address these issues and to make them a priority: failure to adopt policies and regulations on the development and use of facial recognition technology would effectively cede decision-making and rulemaking on these important issues of major public concern entirely to the private sector and the marketplace.

Facial recognition technology has been increasingly incorporated into everyday life, with a wide range of uses, the report states. Some of these uses are innocuous, such as allowing people to unlock their smartphones. But when applied broadly and without safeguards, technology can enable repressive regimes to create detailed records of people’s movements and activities and block citizens’ participation in public life. Many potential uses fall somewhere in between, creating a large grey area where individual assessments of risks, benefits, trade-offs and values may vary, thus affecting how they should be regulated or permitted. The report recognises the value of facial recognition technology and does not advocate a blanket ban, but states that a number of uses may cause sufficient concern to ban them.

The study says that there are two main categories of facial recognition concerns, although they may overlap. One is potential harms from problematic use or misuse of technology, which become more prominent as technology becomes more accurate and capable. The second is potential harms from errors or limitations of the technology itself, such as when systems have different false positive or false negative rates for different demographic groups.

_____

Aquest apunt en català / Esta entrada en español / Post en français

The UK National Statistics Office (ONS) reported a 25% increase in the number of fraud offences in 2021 compared to 2020. Fraud, which accounts for more than 40% of all crimes against individuals, is the most common crime in the United Kingdom.

As if these statistics were not alarming enough, there is the added belief that victims of fraud are often blamed for being foolish or too trusting. But it is time to accept that this can happen to anyone. It has become such a problem that the concept of fraud as an experience suffered only by vulnerable people should be revised. The human brain cannot keep up with all the new types of fraud enabled by technology.

As recently published in The Conversation, a new approach is needed that holds financial institutions and companies accountable for identifying or facilitating fraud and that leverages artificial intelligence to detect suspicious transactions. It is unreasonable to expect consumers to know when they are being scammed if banks and social media platforms cannot.

A 2023 report from UK Finance indicates that there is a growing trend among fraudsters to target people between 18 and 24 years of age, who are much more likely to fall for a phishing scam than people aged 65 and over. In addition, the rate of 13- to 17-year-olds who are victims of video game scams has also risen sharply.

However, the programs on how to protect yourself from fraud that are currently offered are rather scarce. For example, the children’s charity NSPCC has programs to protect minors from online abuse and to keep them safe while using social networks and from possible legal but harmful content, but not to protect them from online scams. Fraud prevention should be taught in schools and universities as part of the curriculum.

One of the most important theories in criminology is deterrence theory, which states that crime reduction is related to the severity of punishment and, more importantly, to the likelihood of being caught. Research suggests that increasing the likelihood of catching the offender is much more effective than increasing the punishment. However, fraudsters have little reason to worry. According to UK government sources, fraud accounts for more than 40% of all crime, but less than 1% of police resources are devoted to combating it.

In fact, the Federal Trade Commission (FTC) noted that a quarter of those who had lost money to fraud said the process had started on social media platforms.

The nature of social media gives fraudsters the ability to hide behind fake profiles and pretend to operate as a legitimate business. They also allow fraudsters to reach millions of people who have only one click to make, among whom younger adults, who tend to make heavier and more prolific use of social networks, are particularly vulnerable.

The FTC has issued orders on a number of social networks, such as Meta, TikTok and YouTube, to request information from them on how they detect malicious ads or scams.

California lawmakers, meanwhile, are considering a bill that would offer seniors greater protection against fraud and hold banks accountable when ATMs facilitate fraudulent transactions.

In the United Kingdom, an anti-fraud strategy was presented to Parliament in May 2023 proposing a number of measures, including a ban on all telephone calls related to financial products.

These two bills are a move in the right direction, but more work needs to be done urgently. Policy makers must allocate funds for research, and police agencies must introduce laws that provide greater protection for individuals and collaborate with international law enforcement agencies such as Interpol.

Fraud affects society at all levels: individuals, organisations and governments.

_____

Aquest apunt en català / Esta entrada en español / Post en français

The metaverse has the potential to change the way we interact and relate to each other and technology. That being said, there are also potential pitfalls and risks, as with any new technology. Potential problems with privacy, security and legislation are part of the downside of the metaverse. This is explained in a recently published report by the website Cointelegraph.

When it comes to metaverse platforms, one of the main problems is privacy. Individuals may share sensitive data and personal information in the metaverse, increasing the risk of hacking and data breaches. In addition, there may be less oversight and regulation of how companies collect and use this data, which could lead to the misuse of their personal data.

Since it is a virtual environment, the metaverse is open to several security risks, such as hacking, intellectual property theft and misuse of user data, which can lead to loss of personal data, financial damage and harm to the reputation and stability of virtual communities. For example, criminals can use the metaverse to commit additional crimes, spread malware or steal personal data.

Regulation is another problem, because the metaverse is a young and rapidly evolving environment. Governments and other institutions may struggle to keep up with technology and lack the resources or tools needed to govern it successfully. This lack of supervision may lead to problems such as illegal activities or dangerous content.

In addition, it is also unclear how the metaverse will affect society, because it is a totally new area that is developing and transforming very quickly. While some experts claim that technology will create more options for community and connection, others respond that it will simply increase social alienation and isolation.

By exploiting flaws in virtual systems and user behaviour, such as malware infections, phishing, and illegal access to personal and financial information, cybercriminals take advantage of the metaverse in a number of ways:

• Phishing: thieves may use phishing techniques to trick victims into revealing personal information or login credentials, which can then be used for identity or data theft or other illegal acts.
• Hacking: to steal money or personal information, criminals may attempt to hack into user accounts or metaverse platforms.
• Malware: to access sensitive data or perform illicit operations, criminals can use malware to infect virtual environments or metaverse-compatible devices.
• Frauds: criminals may take advantage of the anonymity and lax regulation of the metaverse to carry out scams or pyramid schemes.
• Ransomware: thieves may use ransomware to encrypt a user’s digital possessions or personal data before requesting payment in exchange for the decryption key.
• Exploitation of virtual goods and assets: cybercriminals may use bots or other tools to buy virtual goods and assets, which they then sell to the black market for real money.
• Creation of fake digital assets: criminals may create fake virtual assets and sell them to unsuspecting buyers, causing victims to lose money.
• Social engineering: thieves may take advantage of the social elements of the metaverse to gain people’s trust before scamming them.

_____

Aquest apunt en català / Esta entrada en español / Post en français

According to a recent report by CheckPointResearch, hackers are already using the new artificial intelligence chatbot ChatGPT to create new low-level cyber tools, such as malware and encryption scripts. As Sam Sabin reports from the Axios website, security experts have warned that OpenAI’s ChatGPT tool could help cybercriminals accelerate their attacks, and all in a short period of time.

The report lists three cases in which hackers figured out various ways to use ChatGPT to write malicious software, create data encryption tools and write code creating new dark web marketplaces.

Hackers are always looking for ways to save time and speed up their attacks, and ChatGPT’s artificial intelligence-based responses often provide a good starting point for most hackers writing malware and phishing emails.

CheckPoint noted that the data encryption tool created could easily become hijacking software once some minor issues are fixed.

OpenAI has warned on several occasions that ChatGPT is a research preview and that the organisation is constantly looking for ways to improve the product to prevent potential abuse.

The AI-enabled chatbot that has stunned the tech community can also be manipulated to help cybercriminals hone their attack strategies.

The arrival of OpenAI’s ChatGPT tool could allow the fraudsters behind email- and text-based phishing attacks, as well as malware groups, to speed up the development of their schemes.

Several cybersecurity researchers have been able to get the AI-enabled text generator to write phishing emails or even malware for them over the past few weeks.

But it should be clear that hackers were already becoming very adept at incorporating more humane and harder-to-detect tactics into their attacks before ChatGPT came on the scene.

And, often, hackers can gain access through simple computer errors, such as hacking into the corporate account of a former employee still active.

ChatGPT arguably speeds up the hackers’ process by giving them a launching pad, although the responses are not always perfect.

Although OpenAI has implemented some content moderation warnings in the chatbot, it is easy for researchers to circumvent the current system and avoid penalties.

Users still need to have some basic knowledge of coding and attack launching to understand what works correctly in ChatGPT and what needs to be adjusted.

Organisations were already struggling to defend against the most basic attacks, including those in which hackers use a stolen password leaked online to log into accounts. AI-enabled tools such as ChatGPT could only exacerbate the problem.

Therefore, network defenders and IT teams must intensify efforts to detect phishing emails and text messages to stop these types of attacks.

_____

Aquest apunt en català / Esta entrada en español / Post en français

Nearly half of irregular border crossings during 2022 were overland through the Western Balkans region, according to a report by the EU border agency Frontex. Preliminary figures do not include Ukrainian refugees.

The number of irregular border crossings in the European Union increased by 64% last year compared to 2021. According to agency estimates, some 330,000 entries were detected, of which 45% were carried through the Western Balkans region.

The Central Mediterranean route had the second highest number of crossings, increasing by more than half to over 100,000.

Most of the people who attempted the dangerous sea route last year were nationals of Egypt, Tunisia and Bangladesh. Frontex also reports that 2022 reached the highest number of people in five years from Libya, the main departure point from North Africa.

Regardless of the entry route, Syrians, Afghans and Tunisians accounted for approximately 47% of border crossing attempts. The number of Syrians approximately doubled to 94,000.

Males accounted for more than 80% of attempts to enter the Union. The proportion of reported minors decreased by about 9% of all irregular entries.

The latest Frontex figures did not include millions of Ukrainian refugees who entered the EU between February, when Russia invaded Ukraine, and December.

In this vein, during this January, Europol supported Bulgarian authorities during a large-scale day of action against organised crime groups involved in migrant smuggling. The activities, coordinated by the Bulgarian prosecutor and with the participation of the General Directorate for Combating Organised Crime, the National Police and the Border Police, have targeted criminal networks active along the Balkan route. Bulgarian investigations were also coordinated with Turkish and Serbian authorities and other cooperating agencies.

The joint actions took place in Bulgaria and focused on a number of migrant smuggling networks from Turkey through Bulgaria to Serbia and then Western Europe. The main organisers of the networks active along this route are based in Bulgaria, Serbia and Turkey. They have created their own national networks of members responsible for transportation and accommodation in their respective countries.

The main means of transportation used by smugglers were vans, caravans and buses.

Bulgarian authorities have reported an increase in migrant smuggling activities at its southern border. In August 2022, an incident involving a bus carrying irregular migrants resulted in the death of two police officers on duty. Months later the same year, a Bulgarian border police officer was shot dead during a regular patrol on the green border with Turkey. These facts suggest an increase in both smuggling activities and the violence of the criminal networks involved.

_____

Aquest apunt en català / Esta entrada en español / Post en français