SoSafe, the safety awareness and mortal threat management solution, has examined social architectural tactics and the escalating risks that organizations are facing, revealing that nearly nine in ten respondents have experienced some form of intended breach.
The SoSafe report, titled 2025 Cybercrime Trends, examined the opinions of 100 SoSafe users from 10 different nations and 500 global surveillance professionals. AI was a key topic of conversation for the respondents, with 91 % of security experts anticipating a significant rise in AI-driven threats over the next three years.
Only 26 % of respondents express high confidence in their ability to identify these attacks. Given that 87 % of businesses have admitted to having been victims of an AI-driven attack in the last year, it is extremely worrying that only one in four businesses are comfortable in their own resilience.
Over 51 % of security officials cited obfuscation practices, such as AI-generated techniques to mask the roots and intent of attacks, as the top concern, according to the SoSafe report. Additionally, 44 % of respondents said that the development of entirely new attack techniques was their main concern, while 2/5 ( 38 % ) cited the scale and speed of automated attacks.
Advancements in AI are making it possible to launch channel cyberattacks that combine messaging, social media, and partnership platforms. 95 percent of security professionals concur that this type of attack has increased over the past two years. The attack on WWP‘s CEO is an excellent illustration of how the attackers used Microsoft Teams, WhatsApp, and deep fake voice calls created by AI to collect sensitive information and money to create a distinct example.
According to , CSO, SoSafe, “targeting patients across a variety of communication channels makes them appear more legitimate,” Rose said. ” Simplistic email problems are turning into 3D hacking, seamlessly incorporating voice, video, or text-based elements to create advanced schemes with AI.”
Using AI may prevent attacks while also introducing fresh attack options.
In today’s competitive environment, businesses may follow AI because it is essential to meet consumer demand for personalization. It offers a personalized experience, but it can also identify bad actors much more quickly than mortal risk observers. The technology is not illegibly complex, though.
Unintentionally, the use of internal AI makes organizations ‘ attack surfaces more vulnerable, giving rise to novel, creative attacks like AI illusions and data poison. In fact, according to SoSafe’s study, 55 % of companies have not completely implemented controls to control the risks associated with their internal AI options.
Rose continued,” Assaulters can use the benevolent AI that organizations adopt for their own gain to find valuable information, key assets, or pass other controls. Several companies have considered the possibility of their bot becoming an accomplice in an attack by assisting the attacker to gather sensitive data, identify key personnel, and gain valuable corporate insight. Some companies create to offer their staff with assistance.
It is crucial for businesses to integrate a comprehensive approach to security that protects against both modern and individual vulnerabilities.
With education, AI issues don’t outweigh its advantages.
While there are many issues with AI, , CEO of SoSafe, claims that AI is still one of the biggest supporters a company can have in the fight against fraud. However, employees may receive adequate training.
He states,” AI certainly presents new problems, but it also continues to be one of our greatest allies in protecting organizations from ever-evolving threats.” However, the security that is based on AI is only as powerful as the users.
“Cybersecurity knowledge is essential. Perhaps the best technologies falls short when it comes to responding to and recognizing AI-driven risks. We can stay ahead of the curve and create stronger, more resilient organizations by combining people skills, safety knowledge, and the careful application of AI.