GRACE Governance for Large Language Models (LLMs) address and mitigate the concerns and risks associated with the use of LLMs in your organization

COPENHAGEN, DENMARK | APRIL | 2023

2021.AI, a leading AI governance and management software provider, announced today the launch of GRACE Governance for Large Language Models (LLMs). The solution is aimed at addressing and mitigating the risks and concerns associated with the use of LLMs in organizations.

LLMs such as ChatGPT are widely used in daily work in many organizations and businesses. However, many organizations are utilizing AI in a production setting without proper governance in place. This has led to common concerns such as intellectual property leaks to third parties, usage and privacy-related information leaks to third parties, and uncontrolled costs related to the usage of the models.

In response to these concerns, the OECD.AI called for trustworthy AI use cases for tools they have validated to be a part of their global trustworthy AI catalog. GRACE was validated, and our use case, Enterprise ChatGPT, and LLM Governance has been selected for this new catalog as the first solution for LLM Governance to be approved by the OECD.AI.

GRACE Governance for LLMs offers organizations the ability to onboard LLM users according to conformity assessments and company policies. It also masks and anonymizes sensitive information, such as Personal Identifiable Information (PII) to the LLM while still offering to interact with the model. In addition, the solution enables real-time monitoring of the prompting and responses (in- and output from the model), defines guardrails and risks around the LLM, and monitors that they are followed in real-time.

“GRACE Governance for LLMs is a solution that we are proud to launch to help organizations mitigate the risks associated with LLMs,” said Mikael Munck, CEO of 2021.AI. “With the increasing use of AI in organizations, we believe it is important to have proper governance in place, and we are glad to offer a solution that will help organizations achieve this.”

GRACE Governance for LLMs offers a range of benefits, including conformity assessment, real-time LLM monitoring, enhanced security, increased transparency, better control, and operational monitoring. The solution ensures that ChatGPT is being used securely, protecting sensitive information, and preventing data breaches. It also detects responses with biased, discriminatory, or hate language.

To find out more about GRACE Governance for LLMS, please visit 2021.ai/grace-governance-llm or contact CEO Mikael Munck directly by emailing mmu@2021.ai or calling +45 51 595 050.

You might also like…