11/05/2026

LYFE MONDAY | MAY 11, 2026

23

Nearly one in two young people in Europe have used artificial intelligence (AI) chatbots to discuss intimate or personal matters, as the technology increasingly serves as a source of emotional support, an Ipsos BVA survey showed recently. Of the 3,800 people surveyed, 51% said it was “easy” to discuss mental health and personal issues with a chatbot. Only 49% said the same about healthcare professionals and 37% about psychologists. People close to you were at the top of the list, with 68% saying it was easy to discuss issues with friends and 61% with parents. The survey, commissioned by France’s privacy watchdog CNIL and insurer Groupe VYV, was carried out among people aged 11 to 25 across France, Germany, Sweden and Ireland in early 2026. The findings showcased growing concerns over young people’s mental health. About 28% of respondents met the threshold for suspected generalised anxiety disorder, the survey found. Around 90% of those surveyed had used AI M ORE than 50% of leaked passwords end with a number, revealed Kaspersky’s latest research, where its experts analysed 231 million unique passwords in major password leaks from 2023 to 2026 and uncovered several key patterns. First, 68% of modern passwords can be cracked within a day. Second, it turned out that the vast majority of compromised passwords either begin or end with a digit – a common pattern that makes them potentially vulnerable to brute force attacks. Third, users also favour positive and trending words, for example, over the past couple of years, use of the word “Skibidi” in analysed passwords surged 36 times, mirroring the rise of that internet trend. In the recent years, secure passwords’ rules have become a widely discussed topic. More and more services now demand passwords that are at least 10 characters long, include an uppercase letter, and contain a number or a symbol. Yet a comparative analysis of leaked passwords from the past few years shows that that even following some of those rules does not guarantee resistance to brute force or artificial intelligence (AI)-driven attacks. How to make passwords more complex and secure, and how not to repeat common mistakes: Be creative with using symbols and numbers Among the leaked passwords that contain just one symbol, the “@” sign tops the list, appearing in 10% of cases. The next most common symbol is a dot (.), found in 3% of passwords. Among all analysed passwords, “@” takes second place in terms of prevalence and in third place is “!”. Numbers also follow similarly predictable patterns: • 53% of examined passwords end with digits • 17% begin with digits • Nearly 12% include a numeric sequence that resembles a date (from 1950 to 2030) • 3% of leaked passwords include keyboard sequencies such as “qwerty” or “ytrewq”, but most of them are digital sequencies such as “1234” Its data science team lead Alexey Antonov noted that commonly used symbols, numbers or dates – especially when placed in obvious positions (such as at the beginning or end of a password) – significantly simplify brute force attacks for cybercriminals. That is why it is highly recommended to give preference to less popular characters and avoid numeric or keyboard sequences. “Bruteforce works by systematically trying every possible character combination until the correct password is found. When attackers already know which characters users tend to favour, the time required to crack a password drops dramatically. To avoid the temptation of choosing predictable symbols, entrust password creation to dedicated generators

Strengthening passwords o Coming up with something that cannot be easily cracked According to findings, 68% of modern passwords can be cracked within a day.

sued Google, alleging its Gemini AI chatbot contributed to his paranoia and eventual suicide. The results of the survey were not a surprise, said Ludwig Franke Foyen, a psychologist and digital health researcher at Stockholm’s Karolinska Institutet. Current large language models can produce high-quality responses, Foyen said, adding that his research suggested even licensed professionals may struggle to distinguish AI generated advice from that of human experts. But he warned against relying on chatbots alone for mental health support, saying general purpose AI systems were designed for engagement and companies’ goals may not align with mental healthcare needs. “AI can offer information and support, but it should not replace human relationships or professional care. “If someone turns to a chatbot instead of speaking to a parent, a friend or a mental health professional, that is a concern. We do not want technology to make people feel more alone,” said Foyen. – Reuters with the rise of AI-driven tools, length alone no longer guarantees security – even lengthy passwords can be compromised if they follow predictable patterns. The research shows that short passwords of up to eight characters that appeared in the leak are typically cracked by brute force attacks in under a day. However, thanks to AI powered smart algorithms, more than 20% of 15-character passwords can be broken in less than a minute. What is more, 60.2% of all analysed passwords – regardless of length – can be cracked in about an hour. 68.2% – in a day. In the examples provided, the calculations assume a single RTX 5090 GPU and the MD5 algorithm. In real world scenarios, attackers can rent multiple GPUs – 10, 100 or even more. Under such conditions, the cracking rate would increase potentially by several orders of magnitude. In modern terms, truly secure passwords not only meet the gold standard of 16+ characters, but also consist of random, non repeating letters, numbers and symbols, and are unique for each account. For easy and secure password management, auto-fill and cross-device synchronisation, consider using a password manager in which all credentials are stored in a secure vault and protected by a single master password. This eliminates the need to remember hundreds of passwords while keeping them safe from breaches.

How password length affects hack rates: Results based on a single 5090 GPU and MD5 algorithm.

“nightmare” and “scar” also occur. “Using a single-word password, even with a trailing number or a special character, is a weak choice. The pattern is too predictable, making it easy for attackers to guess. Instead, craft a passphrase that strings together several unrelated words, each supplemented with internal numbers and symbols, and sprinkle in a few intentional misspellings. The longer and more random and unpredictable the password is, the harder it is to crack. As an additional way to protect yourself, enable two-factor authentication wherever possible”, recommended Antonov. Is password length important? It is well known that longer passwords are harder to crack and the analysis of leaked passwords confirms this principle. However,

that produce random letters, numbers and symbols with equal probability,” said Antonov. Avoid single-word passwords The research shows that emotional and trending words frequently become the basis for a password. For example, from 2023 to 2026, the use of the word “Skibidi” in passwords increased 36 times – mirroring the rapid rise of that internet trend. Kaspersky experts have also conducted analysis of the occurrence of positive and negative words in passwords, and it turned out that there are more positive ones. Among those regularly appearing are positive words such as “love”, “magic”, “friend”, “team”, “angel”, “star” and “eden”. Interestingly, positive words are much more common than negative ones. However, words such as “hell”, “devil”,

Young people turn to AI chatbots for emotional support, survey shows

Of the 3,800 people surveyed, 51% say it was ‘easy’ to discuss mental health and personal issues with a chatbot. – 123RFPIC

impact of AI tools have also grown over the past year and experts have warned about the limitations of AI in detecting human emotions and safely providing emotional support. Earlier this year, the family of a Florida man

tools before, with many citing their constant availability and non-judgmental nature. More than three in five users described AI as a “life adviser” or a “confidant”. However, concerns over the psychological