Artificial intelligence is transforming access to creativity, but it also poses an increasing risk in the field of cybersecurity. A recent report from the Catalonia Cybersecurity Agency reveals that up to 82.6% of emails with malicious links use generative tools to deceive users. The study "Prospectives 2026," cited by elperiodico.com, warns that the proliferation of generative AI will make cyber scams increasingly sophisticated, automated, and difficult to identify, according to Laura Caballero, the agency's director. Cybercriminals are leveraging the capabilities of large language models, such as ChatGPT, Gemini, Grok, or Claude, to create messages, images, audio, and videos that are nearly indistinguishable from legitimate ones. Although these systems generate content based on statistical data and do not always guarantee accuracy, their realism facilitates deception and identity theft, allowing access to sensitive information, from personal to banking data. This information can be used for ransomware attacks, where files are seized and a ransom is demanded for their release. In Catalonia, cyber scams are already the second most common crime, with about 300 reports daily according to the Mossos d'Esquadra. Credential theft has increased by 160% in 2024, reflecting the expansion of these practices among malicious actors. AI is enabling cybercriminals to scale their operations and increase the effectiveness of spam, Caballero notes. The advancement of AI agents, capable of operating autonomously across multiple applications, adds new risks. Experts like Meredith Whittaker, president of Signal, have warned about the potential dangers of these systems, which can compromise the security of personal and corporate devices. Caballero acknowledges these risks but also highlights the value of AI if implemented with the necessary precautions, and she does not recommend that companies forgo these agents in their processes. Despite the rise in threats, awareness of cybersecurity is also growing. The Generalitat of Catalonia has allocated 18.6 million euros to strengthen the protection of its critical systems, and both the public and private sectors are investing more in prevention and resilience. As a result, the percentage of victims who pay ransoms for ransomware attacks has dropped from 85% in 2019 to 23% in the past year.