AI & Technology

New Study Shows Companies Aren’t Prepared for AI Security Risks

By Allie TribePublished June 6, 2025Updated November 27, 20253 min read
Engineer calling colleagues after finding error in code while doing checkup on server room used for AI workloads
Listen to this article
3 min read

AI use is growing throughout all industries, with 78% of companies worldwide utilizing artificial intelligence. Despite companies’ quick adoption of AI, recent research from BigID, an AI security and data privacy platform, found that most companies’ security measures aren’t up to par for the risks AI brings. 

Published on Wednesday, BigID surveyed 233 compliance, security and data leaders to find that AI adoption is outpacing security readiness, with only 6% of organizations implementing advanced AI security strategies. 

Ranking as the top concerns for companies are AI-powered data leaks, shadow AI and compliance with AI regulations. 

69.5% of organizations identify AI-powered data leaks as their primary concern

As the uses of AI expand, so does the potential for cyberattacks. Increasing amounts of data, from financial records to customer details, and security gaps can make AI systems tempting targets for cybercriminals. The possible consequences due to AI-powered data leaks are widespread, from financial loss to private information breaches, yet according to BigID’s report, nearly half of organizations have no AI-specific security controls.

To help prevent data leaks, BigID recommends regular monitoring of AI systems, as well as who has access to them. Systematic checks for any unusual activity along with implementation of authentication and access controls can help keep AI systems running as designed. 

For an added layer of security, organizations can consider changes for the actual data used in AI. Personal identifiers can be taken out of data or replaced with pseudonyms to keep information private, or synthetic data generation, creating a fake data set that appears exactly like the original, can be used to train AI while keeping an organization’s data safe. 

Nearly half of surveyed organizations worry about shadow AI

Shadow AI is the unmonitored use of AI tools from employees or external vendors. Most often, shadow AI is seen in employee use of generative AI, including commonly used platforms like ChatGPT or Gemini. As AI tools become more accessible, the risk for shadow AI grows, with a 2024 study from LinkedIn and Microsoft showing 75% of knowledge workers use generative AI in their jobs. Unauthorized use of AI tools can lead to data leaks, increased difficulty in regulation compliance and bias or ethical issues. 

The best defense against shadow AI starts with education. Creating clear policies and procedures for AI usage throughout a company, along with regular employee training, can help to protect against shadow AI. 

80% of organizations are not ready or are unsure on how to meet AI regulations

As the uses for AI have grown, so have mandated regulations. Most notably, the EU AI Act and General Data Protection Regulation (GDPR) are the leading European regulations for AI tools and data policies. 

While there are no explicit AI regulations for the U.S. at this time, BigID recommends companies comply with the EU AI Act, enact auditability for AI systems and begin to document decisions made by AI to prepare for more regulations around AI usage. 

As the potential of AI evolves, more companies are prioritizing digital help over human employees. Before your company jumps on the bandwagon, make sure to take the proper steps to safeguard against the new risks AI brings. 

Photo by DC Studio/Shutterstock

Allie Tribe

Allie Tribe

Read Next

What OpenAI’s Rogue AI Hack Teaches About Governance
AI & Technology

What OpenAI’s Rogue AI Hack Teaches About Governance

Your AI agent just found a security hole, taught itself to exploit it, and broke into another company’s servers—without anyone asking it to. If it can happen at OpenAI, it’s already happening at your company.

Read Article →
Alibaba’s New Accio AI Agent Promises to Cut 70% of Manual Workflows for Small Businesses
AI & Technology

Alibaba’s New Accio AI Agent Promises to Cut 70% of Manual Workflows for Small Businesses

Adobe’s New AI Toolkit Takes You From Idea to Ad in Minutes
AI & Technology

Adobe’s New AI Toolkit Takes You From Idea to Ad in Minutes

Zuckerberg Announces Meta Will End Fact-Checking With the Goal of Reducing Censorship
AI & Technology

Zuckerberg Announces Meta Will End Fact-Checking With the Goal of Reducing Censorship

The AI Skills Premium: 62% More Pay, 144% More Demand
AI & Technology

The AI Skills Premium: 62% More Pay, 144% More Demand

Words Worth Carrying

The Quote Collection

Browse all quotes →
Quote by Wyclef Jean
Quote by Cassandra Worthy
Quote by William Nicholson
Quote by Satya Nadella
Save & share these quotes →