Four years after the launch of ChatGPT, generative artificial intelligence has not only transformed the way companies and citizens use technology, but also the capabilities of cybercriminals used in cyberattacks. The automation of certain tasks, the creation of fraudulent content and the sophistication of social engineering techniques are some of the possibilities that this technology has opened up for criminal activity.
This evolution coincides with a significant increase in the activity of cybercriminal groups, whose attack publications have increased by 25% in the first ten months of 2026 compared to the previous year, according to the latest cyber intelligence data compiled by Secure&IT, a company specialized in cybersecurity.
In 2024, 5,613 attacks by cybercriminal groups were published, a figure that rose to 7,979 in 2025, which represents an increase of 42%. This trend has continued in 2026, since between the months of January and October 9,981 publications have been registered, 25% more than during the previous year.
“The month with the most activity in 2026 was August, with 1,368 attacks published, followed by October, with 1,320. Likewise, the analysis that we have carried out in 2025 reflects an increase in the number of cybercriminal groups identified, which went from 114 in 2024 to 162 in 2025. Among the most active, Qilin, Akira and Play stand out, which concentrated more than a quarter of the attacks. publications recorded during that year,” indicates Francisco Valencia, general director of Secure&IT.
Incidents in Spain grew by 46%
This evolution is also reflected in official Spanish statistics. According to the National Cybersecurity Institute (INCIBE), the incidents managed in Spain have increased from 83,517 in 2023 to 122,223 in 2025, an increase of 46%. In that same period, phishing incidents increased from 14,261 to 25,133, which represents a growth of 76%.
Although the increase in cybercriminal activity responds to multiple factors, AI is introducing new capabilities that can facilitate the execution of certain threats. In this sense, organizations such as the European Union Agency for Cybersecurity (ENISA) and Europol have warned that these technologies are helping to reduce certain technical barriers and expand the possibilities of action for cybercriminals.
“This evolution cannot be attributed exclusively to AI, but we are observing how this technology is transforming the way in which cybercriminals operate. The possibility of automating tasks, perfecting deception and accelerating certain phases of attacks poses new challenges for the security of organizations,” explains Valencia.
Artificial intelligence transforms attack techniques
One of the main transformations is taking place in social engineering. Traditionally, many phishing campaigns could be identified by spelling errors, poor translations, or poorly crafted messages. However, generative AI allows you to create more natural communications, adapt them to different languages and personalize them according to the profile of your recipients.
This evolution especially affects fraud directed against organizations, such as the impersonation of managers or suppliers to request financial transfers, obtain credentials or access confidential information. Added to this is the use of synthetic content such as voices, images and videos generated with AI, which can make certain identity theft attempts more credible.
“AI is changing the rules of the game. Attackers can craft much more convincing messages, personalize their deceptions and automate tasks that previously required more time. This poses a significant challenge for organizations, because some of the signals that traditionally allowed them to recognize a threat are no longer reliable,” says the CEO of Secure&IT.
More automation and new capabilities
The use of AI also allows us to streamline the collection and analysis of information, and generate fraudulent content on a larger scale. This transformation coincides with a scenario of growing professionalization of cybercrime, characterized by the existence of specialized groups and business models such as ransomware as a service (RaaS), which allow different actors to access attack tools and capabilities.
“In this context, the incorporation of AI can contribute to reducing certain technical and operational barriers, facilitating the development of some criminal activities,” comments from Secure&IT.
Artificial intelligence offers companies new opportunities to improve their defensive capabilities, facilitating the analysis of large volumes of information, the correlation of security events and the automation of certain responses to incidents.
Companies need to combine technology, processes and specialized knowledge to identify suspicious behavior
Faced with this scenario, organizations must reinforce their protection strategies, combining preventive measures with advanced detection and response capabilities. Training and awareness of the use of AI for employees is essential and must be complemented with access controls, identity verification mechanisms and continuous monitoring to identify suspicious behavior.
“The challenge is no longer protecting systems against a greater volume of threats, but rather anticipating attacks that are constantly evolving. Companies need to combine technology, processes and specialized knowledge to identify suspicious behavior, respond quickly and minimize their impact. AI must also become an ally of defense,” concludes Francisco Valencia.
