PERSUIT uses artificial intelligence (AI) to optimize legal services procurement, support decision-making, and improve operational efficiency. The AI capabilities in the PERSUIT platform help you streamline processes, manage costs, and automate workflows.
Technology that powers AI in the platform
The PERSUIT platform uses a combination of AI technologies.
Generative AI
The PERSUIT Outside Counsel Management platform uses Google Foundation Models and approved Partner Models via Vertex AI Model Garden as the core Generative AI technology. The PERSUIT eBilling platform uses Anthropic Claude models via AWS Bedrock. For details on these models, see Google models, Partner models, Vertex AI Model Garden, and AWS Bedrock Models.
PERSUIT supported models:
- Via Google Vertex AI Model Garden:
- Google Gemini 2.5 and above
- Anthropic Claude Haiku 3.5 and above
- Anthropic Claude Sonnet 4.6 and above
- Anthropic Claude Opus 4.6 and above
- Via AWS Bedrock (PERSUIT eBilling platform):
- Anthropic Claude Haiku 3.5 and above
- Anthropic Claude Sonnet 4.6 and above
- Anthropic Claude Opus 4.6 and above
Statistical AI
PERSUIT uses proprietary statistical models, including percentiles, quartile coefficients, and cosine similarity algorithms, to analyze data and provide insights.
MCP Access
Persi, PERSUIT's outside counsel management agent, is also accessible through the Model Context Protocol (MCP). MCP is an optional, opt-in access channel for customer-approved AI assistants, such as Anthropic Claude, ChatGPT, Microsoft Copilot, Google Gemini, or any other MCP-enabled AI assistant.
Access is enabled per organization and authenticated through OAuth or single sign-on. The same authentication and authorization framework that applies to Persi governs MCP access, so you can only access what your PERSUIT permissions allow. For setup steps, see Connect Persi to Claude, ChatGPT or Copilot, or any MCP-enabled AI.
These technologies work together to provide assistance throughout your legal operations.
How PERSUIT protects your data
The security and privacy of your data are a priority. The AI implementation in the PERSUIT platform includes measures to keep your information protected and confidential.
Data Privacy & Compliance
PERSUIT adheres to GDPR and other relevant data privacy regulations, processing all personal data to the highest legal standards.
Secure Processing
When AI processes your data, it remains securely within PERSUIT's cloud infrastructure (Google Cloud Platform (GCP) for the PERSUIT Outside Counsel Management platform and Amazon Web Services (AWS) for the PERSUIT eBilling platform), never leaving PERSUIT's controlled environment.
MCP Access via Customer-Approved AI Assistants
When you access Persi through MCP, PERSUIT's authentication, authorization, and permission controls apply to every request, and each user connects individually with their own PERSUIT account.
Data returned to your AI assistant is processed by that assistant under your organization's own agreement with its provider. The retention, training, and security commitments described in this article for Google Vertex AI and AWS Bedrock do not extend to third-party AI assistants you choose to connect.
Important: Check that any AI assistant you connect to PERSUIT meets your organization's security and data governance requirements.
Stateless AI Models
PERSUIT AI models are designed to be stateless, meaning they do not retain customer data after processing. Each interaction is isolated, and data is immediately discarded.
No AI Training on Customer Data
PERSUIT AI models are never trained on your proprietary or confidential customer data, so your information is not used to improve or modify the AI models themselves.
Encryption
All your data is encrypted using industry-standard methods: AES-256 for data at rest and TLS 1.2 or higher for data in transit.
Secure Cloud Hosting
PERSUIT's infrastructure is hosted on Google Cloud Platform (GCP) and Amazon Web Services (AWS), both of which are certified under ISO 27001 and SOC 2 Type 2, providing a secure and compliant environment for all data processing.
Anonymized Data for Statistics
Statistical analyses and recommendations use fully stateless, aggregated, and anonymized data, which protects privacy while providing useful insights.
Data Retention
As per the PERSUIT enterprise agreement, neither Google Vertex AI nor AWS Bedrock retains any data that PERSUIT shares with it. Zero Data Retention applies at the AI model layer across both platforms. Application-layer retention specific to Persi is described under Tracking of AI Usage.
Tracking of AI Usage
PERSUIT monitors the performance of its AI features as part of its standard software quality and incident management processes. For AI features other than Persi, PERSUIT does not log the prompts or AI generated outputs.
A small internal AI team at PERSUIT monitors Persi interaction logs for quality assurance purposes only, and permanently deletes all interaction logs after 30 days. Zero Data Retention still applies at the AI model layer. Persi is powered by Anthropic and Gemini models accessed exclusively through Google Vertex AI or AWS Bedrock, where prompts and outputs are not retained by the models and no data is used for model training.
PERSUIT also tracks low level logs within PERSUIT's secure cloud environment, capturing metadata such as user ID, organization ID, and timestamps. PERSUIT continuously evaluates model performance to maintain reliability and accuracy across supported features.
PERSUIT's commitment to responsible AI
PERSUIT is committed to the responsible and ethical use of AI. Internal policies and practices ensure that AI augments human capabilities, not replaces them, with a focus on fairness, transparency, and accountability:
- Ethical Deployment: PERSUIT uses AI systems responsibly and avoids actions that could violate privacy or facilitate malicious activities.
- Compliance: All AI usage adheres to applicable laws and regulations, including those concerning data protection, privacy, and intellectual property.
- Transparency & Accountability: PERSUIT maintains transparency about AI involvement in processes and remains accountable for outcomes generated by AI systems.
- Secure Data Handling: PERSUIT follows its privacy and security policies for all data used by AI, so sensitive information is handled with care.
- Bias Mitigation: PERSUIT works to identify and mitigate potential biases in AI systems to support fairness and inclusivity.
- Human Oversight: AI has limitations, so human judgment remains central, especially when you interpret and act on AI-generated recommendations. When Persi is accessed through MCP, AI assistants can query, analyze, and start draft requests. Reviewing, inviting firms, and publishing always take place within the PERSUIT platform.
PERSUIT implements AI with careful consideration for its impact and prioritizes the security, privacy, and trust of its users.
Related articles
- Introducing Persi, your outside counsel management agent
- Connect Persi to Claude, ChatGPT or Copilot, or any MCP-enabled AI
If you need more help, contact the Support team at support@persuit.com.
Comments
0 comments
Article is closed for comments.