How to beat ‘shadow AI’ across your organization

[ad_1]

GenAI is the most disruptive technology to hit society since the internet. Two years on from the launch of the most popular Large Language Model, ChatGPT, GenAI tools have fundamentally and forever changed the way we consume information, create content and interpret data.

Since then, the breakneck speed at which AI tools have emerged and evolved has meant that many businesses have found themselves on the back foot when it comes to the regulation, management and governance of GenAI.

This environment has allowed ‘Shadow AI’ to run rampant. According to Microsoft, 78% of knowledge workers regularly use their own AI tools to complete work, yet a huge 52% don’t disclose this to employers. As a result, companies are exposed to a myriad of risks, including data breaches, compliance violations, and security threats.

Addressing these challenges requires a multi-faceted approach, comprising of strong governance, clear communication, and versatile monitoring and management of AI tools, all without compromising on staff freedom and flexibility.

Adam Wignall

General Manager at Kolekti.

Table of Contents

Trust is paramount, and goes both ways

Employees will use GenAI tools, whether their employer mandates it or not. In fact, blanket bans, or stringent restrictions on how it should be used, is only likely to exacerbate the challenge of ‘Shadow AI’. A recent study even showed that 46% of employees would refuse to give up AI tools, even if they were banned.

GenAI is an incredibly accessible technology which has the power to significantly enhance efficiencies and bridge skills gaps. These transformative tools are within arms reach of time-pressured staffers and employers cannot, without reasonable justification, tell them they’re not allowed to use it.

Thus, the first step for employers to strike the right balance between efficiency and authenticity is to establish the blueprints for how GenAI can, and should, be used within a business setting.

Comprehensive training is therefore essential to ensure employees know how to safely and ethically use AI tools.

This goes beyond technical know-how – it also includes educating staff on the potential risks associated with AI tools, such as privacy concerns, intellectual property issues, and compliance with regulations like GDPR.

Clearly explaining these risks will go a long way in getting staffers on board with those restrictions which may, at first, seem too severe.

Outline clear use cases

Defining clear use cases for AI within a given organization is also extremely important, not just for telling employees how they can’t use AI, but also how they can use it. A recent study actually found that a fifth of staff don’t use AI currently because they don’t know how to.

Thus, with the right training, awareness, and understanding of how AI tools can be used, they can avoid unnecessary experimentation that may expose their organization to risk, while also reaping the efficiency rewards that naturally come with AI.

Of course, clear guidelines should be set around what AI tools are acceptable for use. This may differ depending on departments and workflows, so it’s important that organizations adopt a flexible approach to AI governance.

Once use cases are defined, it’s critical to measure AI’s performance precisely. This includes setting benchmarks for how AI tools are integrated into the daily workflow, tracking productivity improvements, and ensuring alignment with business goals. By establishing metrics to monitor success, businesses can better track the adoption of AI tools, ensuring that they are not only used effectively but that their usage aligns with organizational objectives.

Addressing BYO-AI

One of the main reasons Shadow AI festers, is that employees can bypass IT departments and implement their own solutions through unsanctioned AI tools. The decentralized, plug-and-play nature of many AI platforms allows employees to easily integrate AI into their daily work routines, leading to a proliferation of shadow tools that may not adhere to corporate policies or security standards.

The solution to this problem is through versatile API management. By implementing robust API management procedures, organizations can effectively manage how internal and external AI tools are integrated into their systems.

From a security perspective, API management enables businesses to regulate access to data, monitor interactions between systems, and ensure that AI applications are only interacting with the appropriate datasets in a controlled and secure manner.

However, it’s important to not cross the line into workplace surveillance by monitoring specific inputs and outputs from business sanctioned tools. This is only likely to force AI users back into the shadows.

A good middle ground is for sensitive alerts to be configured to prevent accidental leaks of confidential data. For example, AI tools can be set up to detect when personal data, financial details, or other proprietary information is being input or processed by AI models inappropriately. Real-time alerts provide an additional layer of protection, ensuring that breaches are identified and mitigated before they escalate into full-blown security incidents.

A well-executed API strategy makes it possible to provide employees with the freedom to use GenAI tools productively, while simultaneously safeguarding the originations’ data and ensuring that AI usage complies with internal governance policies. This balance can drive innovation and productivity without compromising security or control.

Striking the right balance

By establishing strong governance with defined use cases, leveraging versatile API management for smooth integration, and continuously monitoring AI usage for compliance and security risks, organizations can strike the right balance between productivity and protection. This approach will allow businesses to embrace the power of AI while minimizing the risks of ‘Shadow AI’, ensuring that GenAI is used in ways that are secure, efficient, and compliant while allowing them to unlock crucial value and return on investment.

We’ve compiled a list of the best network monitoring tools.

This article was produced as part of TechRadarPro’s Expert Insights channel where we feature the best and brightest minds in the technology industry today. The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/news/submit-your-story-to-techradar-pro

[ad_2]