The founders of Grizzly AI tackled AI Security by building a safe-for-work access path for risk-averse enterprises to leverage the amazing power of generative AI.
Grizzly AI offers enterprise-ready automations that integrate with any large language model (LLM) via a secure API—including OpenAI’s GPT model family, Claude, Mistral, and more.
Grizzly is ISO 27001 and ISO 9001 certified, and provides flexible deployment options to match your unique security policies and requirements.
You always retain ownership of your data and any AI-generated content, and built-in guardrails ensure your private information is never used to train external AI models.
In short, Grizzly AI gives you complete control and confidence in harnessing generative AI for automation—while safeguarding your data and keeping you firmly in the driver’s seat.
Below is a summary of Grizzly AI’s key security features.
What security controls does Grizzly AI have?
Software penetration-testing
Grizzly AI’s software is penetration-tested annually by the cyber security team of a Big Four firm, together with Grizzly AI’s compliance with other key industry cyber security standards.
ISO27001 and ISO9001 certification
Grizzly AI Limited is both ISO 27001 and ISO 9001 certified, ensuring that your data meets the highest international standards for information security.
Data movement
The Grizzly AI platform breaks documents or questions up into subsets, or ‘tokens’, to enable the generative AI process to work.
‘Tokens’ are approximately four characters long and are sent to model provider servers for processing.
Grizzly AI offers flexible hosting and server location options—including country-level selection and choice of cloud provider—to align with local data governance and compliance requirements.
Any ‘tokens’ sent to large language models are held for a maximum of 30 days, to enable abuse monitoring to be maintained.
This data is not viewed by any employees of the model provider, or used to train any of their underlying models.
Encryption of data in motion
Client data uploaded to Grizzly AI and the AI responses are all encrypted in motion.
The ‘tokens’ sent by Grizzly AI to model providers and the responses received are all encrypted in motion.
Location of client data
Any company file uploaded to Grizzly AI is protected in a secure environment in the country of your choosing—whether in Microsoft Azure or another compliant infrastructure, depending on your organization’s requirements.
Grizzly AI supports a variety of large language models, while ensuring that none of your uploaded files are ever used for LLM training.
'Token' data
Whenever a question is asked of a file or folder within a client’s knowledge base within Grizzly AI, ‘tokens’, or a subset of this data, is sent to your chosen model provider, where the AI is executed.
Drastic reduction in potential 'hallucinations'
When any generative AI analysis is done on company documents loaded into Grizzly AI, only those company documents are used as a source for generating any outputs. This drastically reduces the potential for ‘hallucinations’.
This is because Grizzly AI is only using the natural language generation (NLG) capability and ‘human’ reasoning of large language models to analyse the information, rather than seeking answers outside the scope of the company documents.
In addition, any results of that analysis also provide references to the paragraph, page and source document, thus allowing easy verification that the results are not ‘hallucinations’.
Option to restrict user access to company documents only
Company administrators of Grizzly AI can, if they wish, prevent any or all users from accessing the wider functionality of large language models. This restriction significantly reduces the possibility of ‘hallucinations’.
That is, the employee is restricted to only company generative AI tasks using only the company’s own documents. This feature, together with other ‘guardrails’, or controls, helps to ensure that enterprise data is protected.
Option for automatic or user-driven deletion of files
Any files uploaded to Grizzly AI’s repositories may be automatically deleted in a regular cadence or may be deleted at will by the authorized user.
Protecting company 'prompts'
The ‘prompts’, or user questions of the generative AI, must also be closely guarded. OpenAI’s, Anthropic's, and other model providers legal partnership agreement with Grizzly AI ensures that ‘prompts’, or any corporate data, will not be used by them for any training of large language models.
Hedge against different generative AI models
OpenAI’s LLM's are widely popular, at present. However, it is very likely that there will be no one generative AI model to rule them all. Exponential progress in generative AI means that winners and losers will change places over time.
Google’s Gemini, Anthropic's Claude, and many open-source upstarts will vie for supremacy. Specialist, or vertical market, generative AI models will rapidly emerge.
Grizzly AI is model-agnostic, enabling seamless integration with multiple generative AI models. They offer expert guidance to help identify the best model for any given workflow, ensuring you remain adaptable and can easily switch providers as technology evolves—without disrupting user experience or existing integrations.
This means that enterprise customers are protected from shifts in the winners and losers of the generative AI technology race, while preserving user skills and familiarity with Grizzly AI, together with most investment in integrations.
Use the best generative AI model for the purpose required
Grizzly AI’s platform enables customers to choose the ‘best’ generative AI models to use for their purpose or multiple models could be used for different purposes.
A customer could, for example, use a highly secure generative AI model for some applications and use other models that might suit specialised requirements.
This allows businesses to capitalise on the benefit of using multiple generative AI vendor models, to obtain the best mix and match of security and functionality to maximise their benefits.
'Walled-garden' access to generative AI
Grizzly AI has created a walled-garden generative AI solution in partnership with leading model providers, including OpenAI. Because firms use their own high-quality information and data, the results are accurate and far superior to a more general use of generative AI models.
The curse of potential ‘hallucinations’ is greatly reduced, primarily because Grizzly is only leveraging the company data and documents uploaded to the Grizzly AI application.
