In 2024, more than 627,000 incidents of breaches in ICT system security were reported in Poland. This represents an increase of around 60% year on year, compared with nearly 391,000 cases recorded in 2023. Cybercriminals are using increasingly sophisticated attack techniques, including those powered by artificial intelligence. Recently, particular attention has been drawn to so-called vibe hacking, which—due to its high level of automation and speed—may pose a serious threat to cybersecurity.
“Vibe hacking is the latest cyberattack technique. Put simply, it no longer requires advanced coding skills—only prompting,” says Ewelina Wachulec, Senior Engineering Manager at Dell Technologies, in an interview with Newseria. “Cybersecurity—and above all cybercrime—is becoming democratized. What once required years of learning and interdisciplinary teams can now be done by a single person using artificial intelligence, which generates malicious code and much more on their behalf.”
According to the expert, this fundamentally changes the scale and pace of cyberattacks.
“Artificial intelligence automates the entire attack process, from start to finish. That means we are more exposed to the effects—campaigns are more likely to succeed, sensitive data may be stolen, or victims may face ransom demands,” Wachulec explains.
AI as an active operator in cybercrime
The growing role of AI in cybercrime is highlighted in Anthropic’s Threat Intelligence Report: August 2025. Researchers identified an operation labeled GTG-2002, marking a new phase in the use of artificial intelligence—particularly coding agents—in cyberattacks, referred to as vibe hacking. In practice, this involves instructing AI to generate code designed to carry out attacks or bypass security measures.
In the analyzed case, attackers used Anthropic’s coding tool to automate both the preparation and execution of attacks. Within a single month, 17 organizations were targeted, including entities in the government, healthcare, emergency services, and religious institutions. Analysts concluded that AI acted not only as an adviser but also as an active operator.
“There was a case in July 2025 where a widely available model was used to hack 17 organizations and steal confidential data,” Wachulec notes. “The good news is that these models are being secured, and there is strong cooperation between technology providers, security services, and cybersecurity teams. In my view, it is no longer possible to broadly use publicly available models for vibe hacking and cyberattacks—whereas not long ago, it was. This is a race: hackers’ methods are evolving, but defensive methods are evolving as well.”
Deepfakes and AI-driven fraud on the rise
Artificial intelligence is also being used in other types of cybercrime. According to the Government Plenipotentiary for Cybersecurity’s 2024 report, Poland’s cyberspace is expected to see an increase in campaigns during 2025–2026 that use AI to generate fake voices and images.
Research titled “Poles’ Attitudes Toward Cybersecurity 2025”, conducted by the Warsaw Institute of Banking together with the Polish Bank Association, shows that 23% of Poles consider deepfakes to be the greatest online threat. Such techniques are commonly used in investment scams, including those related to cryptocurrencies.
“We don’t have precise data on who is most often targeted by AI-driven attacks,” says the Dell Technologies expert. “What’s interesting is that AI does not make decisions the way humans do. It simply and automatically searches for vulnerabilities. It doesn’t care who the victim is—it relentlessly scans for weaknesses, prepares attack scenarios, and launches them.”
Businesses under growing pressure
The scale of cyber threats is also affecting Polish companies. According to KPMG’s Cybersecurity Barometer, 83% of Polish firms recorded at least one cybersecurity incident in the past year—16 percentage points more than a year earlier.
As a result, businesses are placing greater emphasis on digital security. The use of AI to support IT departments ranks second among the most common applications of artificial intelligence in companies, cited by 34% of organizations using AI. However, only 13% say they use AI directly in cybersecurity. According to the report’s authors, this may be due to staff shortages and still-limited financial resources, despite steadily rising investment levels.
“I believe companies do have tools to defend themselves against cybercrime, because cybersecurity is not a new issue,” Wachulec says. “What has changed is the scale of attacks and access to them. This is an ongoing race, because both sides have access to the same technology. What we are clearly lacking, however, is education and awareness of these threats.”
Education remains the weakest link
The report Cyber Portrait of Polish Business 2025, prepared by ESET and DAGMA Bezpieczeństwo IT, shows that only one in five employees rates their cybersecurity skills highly. At the same time, one in three respondents admits to having believed a deepfake they encountered.
“Understanding cyberattack patterns is essential—otherwise we won’t be able to defend ourselves,” Wachulec emphasizes. “If this knowledge is supported by effective detection mechanisms, attacks can be blocked. We also need to be aware that while public models have safeguards preventing misuse, there is an underground ecosystem where models are developed specifically for malicious purposes.”
Education is needed both in organizations and at home. Around 30% of Poles believe that teachers should be responsible for shaping young people’s awareness of safe and healthy internet use through mandatory digital education classes, according to the Poles’ Attitudes Toward Cybersecurity 2025 study. Another 28% argue that children should learn proper digital habits at home, through example and a balanced approach to offline and online life.
“We need education at every level,” Wachulec stresses. “Schools should teach what deepfakes and voice cloning are—that a voice heard on the other side of the screen is not necessarily a well-intentioned human. This knowledge should not be confined to cybersecurity teams alone. It needs to be widely shared—in companies, in schools, and across society.”
Even when awareness of current threats is relatively high, it must be constantly updated. As the expert concludes, the emergence of vibe hacking shows just how quickly new forms of cyber threats can appear.
