Skip to content

Latest
Related Sites
About
Media Kit
Search for:

AI at the Edge
AI Careers
Building with AI
Conversational AI
Cybersecurity AI
Data Engineering
Data Science
More

Home » model safety report

model safety report

Lacking AI Safety Net, Some Powerful Models Come With Trust Issues

By Jon Swartz|2025-04-21T17:54:08-04:00April 21, 2025|

Home
About
Meet our Authors
Write for Techstrong AI
Media Kit
Sponsor Info
Copyright
Terms of Use
Privacy Policy

FOLLOW US

NEWSLETTER SIGN UP

© Techstrong Group, Inc. ALL RIGHTS RESERVED.

Security in AI

Step 1 of 7

14%

How would you best describe your organization's current stage of securing the use of generative AI in your applications?(Required)

We don’t have or use generative AI in our applications

Nothing significant yet - We have or will have generative AI in our applications but haven’t yet researched or explored the possibility of securing it.

Starting - We have begun to work on securing generative AI in our applications.

Developing - We have started to implement some security measures for generative AI.

Average - We have some policies or tools in place.

Advanced - We have some enhanced security protocols specifically tailored for generative AI in our applications.

Very advanced - We have comprehensive, mature, and proactive security solutions for generative AI in our applications.

Have you implemented, or are you planning to implement, zero trust security for the AI your organization uses or develops?(Required)

Already implemented

Planning to implement

Considering or scoping to implement

Not considering zero trust for AI

Do not use zero trust at all

Don’t know

What are the three biggest challenges your organization faces when integrating generative AI into applications or workflows? (Select up to three)(Required)

Integration capabilities of current systems

Access security

Data security or privacy

Lack of expertise or experience

Compliance with standards and regulations

Security of prompts or responses

Cost to implement

Potential impacts on performance or scalability of the applications

Potential impacts on performance or scalability of the AI

Intellectual property protection

How does your organization secure proprietary information used in AI training, tuning, or retrieval-augmented generation (RAG)? (Select all that apply)(Required)

Encryption at rest

Encryption in transit

Data access control, monitoring

Data anonymization

Data/information isolation

Data masking/tokenization

Data governance policies

Zero trust / continuous authentication and authorization

No specific measures in place

Which of the following kinds of tools are you currently using to secure your organization’s use of generative AI? (select all that apply)(Required)

Native cloud service tools (AWS Guardrails, e.g.)

Third-party tools

Custom in-house tools

No specific tools

How valuable do you think it would it be to have a solution that classifies and quantifies risks associated with generative AI tools?(Required)

Extremely valuable

Very valuable

Somewhat valuable

Not valuable

What are, or do you think would be, the most important reasons for implementing generative AI security measures? (Select up to three)(Required)

Regulatory compliance

Protecting intellectual property

Mitigating potential data breaches

Maintaining customer trust

Competitive advantage

Improved risk management

Increased operational efficiency

Don’t know