Financial institutions have been told artificial intelligence implementation is creating new pathways for cyber criminals.

Researchers at Netskope Threat Labs have tracked the key cyber threats facing financial services organisations and employees and released their latest findings in their annual Financial Services report.

It explained that financial services organisations rapidly adopt generative AI (genAI), the risk of exposing sensitive financial and customer data is increasing. Regulated data accounts for 59% of all data policy violations related to genAI usage, highlighting the scale of the challenge in protecting compliance-sensitive information. Intellectual property (20%), source code (11%), and passwords and API keys (9%) also contribute to exposure risks.

“These risks are particularly significant given how widespread genAI usage has become across the financial services sector, with 70% of users actively using genAI tools and 97% interacting with applications that incorporate genAI-powered features indirectly,” the report added. “In addition, 94% of users are using genAI applications that rely on user data for training.

“At the same time, organisations are making progress in reducing shadow AI usage. The proportion of users relying on personal genAI applications has dropped significantly from 76% to 36% over the past year, while adoption of organisation-managed genAI solutions has increased from 33% to 79%. However, the number of users switching between personal and organisation-managed genAI accounts has risen from 9% to 15%, increasing the risk of sensitive financial data moving between unmanaged and secure environments.”

It added that the genAI ecosystem is also diversifying. ChatGPT remains the most widely used application, adopted by 76% of organisations, followed by Google Gemini at 68%. Newer tools are also gaining traction quickly, with Google NotebookLM reaching 39% adoption and AssemblyAI rising sharply from just 1% in June 2025 to 37%, reflecting growing demand for specialised AI capabilities.

At the same time, organisations are taking a cautious approach to risk, with tools such as ZeroGPT (46%), DeepSeek (44%) and PolitePost (43%) among the most frequently blocked genAI applications due to security and compliance concerns.

Beyond AI, the use of personal cloud and online applications in the workplace continues to create data security risks. Regulated data accounts for 65% of data policy violations in these personal apps, showing that sensitive financial information is most exposed when employees operate outside managed environments. Across the financial services sector, LinkedIn (92%), Google Drive (84%), and ChatGPT (77%) are among the most widely used personal apps in the workplace.

The study said adding to these risks, attackers are frequently exploiting trusted cloud platforms to distribute malware. GitHub is now the most abused platform for malware, impacting 11% of organisations, followed by Microsoft OneDrive at 8%. By using trusted cloud infrastructure rather than suspicious domains, attackers are able to blend malicious activity into normal cloud traffic, making threats significantly harder to detect.

Ray Canzanese, director of Netskope Threat Labs, said: “As financial institutions accelerate their adoption of generative AI, they are also expanding the number of pathways through which sensitive data can be exposed. While the shift towards organisation-managed tools is a positive step, our findings show that risks persist, particularly where personal and enterprise usage overlap. To reduce risk, organisations need a layered approach – inspecting all web and cloud traffic to stop malware, blocking non-essential applications, and using data loss prevention to protect sensitive information. Technologies like remote browser isolation also play a key role in enabling safe access to higher-risk websites.”