As digital surface area expands, human error remains a core cybersecurity risk
As digital surface area expands, human error remains a core cybersecurity risk
Publish Date: 2026-03-09 09:41:00
Source Domain: www.wealthprofessional.ca
Using an unordered list, summarize the following article with between 4 and 8 key points.
AI amplifies the risks now facing financial institutions. By lowering the barrier to entry for more sophisticated forms of attacks, as well as opening the door to more effective mimicry of people’s faces and voices. Even in the way it allows for easier digestion of data, generative AI tools are making scammers’ lives that much easier.
Adhering to regulation remains an important part of meeting that challenge. Curliss stresses his firm’s own adherence to SOC2 compliance regulations to ensure they’re meeting or exceeding industry standards for data controls. That also means constant updating of systems, auditing processes, testing defenses with third-party penetration checks, and constantly training people to adhere to these standards.
“People are under tremendous competitive pressure to perform at a very high level, be more efficient, and provide better quality of service,” adds Alexandre Ackermans, VP of Product & Engineering at Maximizer. “If you use ChatGPT, it’s very easy to copy‑paste something from somewhere because you don’t have an AI‑approved tool that is as convenient and then paste that in some other application. That’s an example of where that data is going to move outside of the trusted system and into a system that is less controlled.”
Ackermans sees countless easy mistakes emerging from this impulse towards convenience and ease. Often they occur when a large institution hasn’t been able to roll out an equivalent tool with appropriate data controls. They can happen through small third-party apps that someone might put on their phone to read data from photos, or even through online sources where sensitive parts of the internal security process are shared for public consumption. Just as people within organizations need to be trained and coached to avoid these mistakes, Ackermans and Curliss also emphasize the importance of building secure systems that people would rather use.
Removing friction from existing systems, while maintaining security, is key to this approach. That means building a great UI and UX, as well as ensuring you have your own secure AI system to replace that impulse to just throw a question into public ChatGPT. Curliss notes that Maximizer has done exactly this with their own AI system.
