CoCounsel AI

Description

CoCounsel, developed by Thomson Reuters, is an industry-leading AI platform engineered for professionals who cannot afford to be wrong. It goes beyond mere speed enhancement, aiming to improve the quality and defensibility of professional work. CoCounsel is built on the principle of Fiduciary-Grade AI™, a standard for AI in high-stakes professions that emphasizes duties of care and regulatory oversight.

The platform is designed to accelerate legal, tax, audit, compliance, and drafting tasks with purpose-built AI. Its core strength lies in being grounded in authority, transparency, and trust. Every experience is crafted to help professionals produce work that can be examined, explained, and defended, ensuring it stands up to scrutiny. CoCounsel achieves this by being grounded in authoritative, expert-validated content and proprietary sources.

Integration is a key aspect of CoCounsel's design, connecting seamlessly with tools professionals already use, including Westlaw, Practical Law, Checkpoint, Microsoft 365, and various document management systems. This frictionless integration ensures that users can leverage AI without disrupting their existing workflows. Furthermore, CoCounsel is protected by enterprise-grade privacy and security safeguards, ensuring that sensitive professional work remains confidential and protected.

CoCounsel differentiates itself from generic AI tools by providing answers grounded in authoritative content and offering transparency and verifiability. When accuracy is paramount, CoCounsel surfaces sources and reasoning, allowing users to confidently verify and defend their work. This domain expertise, combined with rigorous testing and refinement by subject-matter experts, ensures the AI's accuracy and nuance for professional workflows.

Trusted by over 1 million users, CoCounsel is not intended to replace professional judgment but to support it. Professionals remain ultimately responsible for their advice, decisions, and signed-off work. The platform's commitment to enterprise-grade protection means customer data is secured with rigorous privacy and security controls, including TLS 1.2 encryption in transit and AES-256 encryption at rest. All GenAI LLM requests are processed under the identity of Thomson Reuters, never identifying the specific customer, and interactions with AI providers use zero-retention API calls.

CoCounsel AI's Core Features

• Fiduciary-Grade AI™ for high-stakes professional work

• Accelerates legal, tax, audit, compliance, and drafting tasks

• Grounded in authoritative content and expert-validated sources

• Integrates with Westlaw, Practical Law, Checkpoint, Microsoft 365, and DMS partners

• Enterprise-grade privacy and security safeguards

• Transparent and verifiable outputs with surfaced sources and reasoning

• Designed to support, not replace, professional judgment

• Zero-retention API calls for AI model interactions

• Data secured via TLS 1.2 and AES-256 encryption

• User content and prompts not used for AI model training

• Customizable data retention policies

• Prompting guardrails to restrict responses to verified information

How to use CoCounsel AI?

1. 1

   Configure: Connect CoCounsel to your existing tools and systems.

2. 2

   Input: Provide your professional documents or queries.

3. 3

   Generate: Utilize CoCounsel's AI to accelerate drafting and analysis.

4. 4

   Verify: Review surfaced sources and reasoning for defensible outputs.

5. 5

   Integrate: Seamlessly incorporate AI-assisted work into your workflow.

6. 6

   Optimize: Leverage CoCounsel to enhance efficiency and accuracy.

CoCounsel AI's Use Cases

• Legal Document Drafting
• Tax Research and Analysis
• Audit Work Acceleration
• Compliance Work
• Data Management
• Client Service Enhancement
• Work Scrutiny and Defense

FAQ from CoCounsel AI

• Does Thomson Reuters retain user content or prompts for AI training?

  No, your user content and prompts are not used to train or improve CoCounsel or associated products and LLMs. Thomson Reuters AI third-party partners are contractually prohibited from using customer data for training, and systemic controls are in place to prevent human access to customer data.

• How long does Thomson Reuters retain user content and prompts?

  Retention is customized at the organization account level with a Thomson Reuters client manager and deleted according to the customer's policy. Users can also delete their content and prompts within the product at any time.

• How does Thomson Reuters protect user privacy and data?

  All GenAI LLM requests are processed under the identity 'Thomson Reuters.' Data is secured in transit via TLS 1.2 and at rest via AES-256 encryption. Interactions with GenAI LLM providers use zero-retention API calls.

• Where is user content and prompts stored?

  Systematic procedures and hosting are in the U.S. All interactions with OpenAI GPT and Google Gemini are processed in the U.S. Backup data is hosted in separate availability time zones in the U.S., with region-based hosting available in the UK, Australia, and Canada.

• What guardrails prevent bias in Thomson Reuters content?

  Thomson Reuters has implemented prompting guardrails ensuring GenAI products produce responses only from a defined and limited set of verified information, such as their own content databases, restricting the LLM's ability to source broadly from the internet.

• What is Fiduciary-Grade AI?

  Fiduciary-Grade AI is Thomson Reuters' standard for AI in high-stakes professions. It is AI designed for professionals with duties of care and regulatory oversight, drawing on authoritative content, protected by privacy and security safeguards, shaped by experts, and producing transparent, verifiable outputs.

• How is Fiduciary-Grade AI different from general-purpose AI?

  General-purpose AI is for broad productivity tasks. Fiduciary-Grade AI is specifically designed for professional work where outputs influence legal judgments, financial disclosures, regulatory filings, or client advice, emphasizing authoritative sources, privacy, expert involvement, and verifiable outputs.

• Does Fiduciary-Grade AI replace professional judgment?

  No, Fiduciary-Grade AI is designed to support human judgment, not replace it. Professionals remain responsible for the advice they give, the decisions they make, and the work they sign off on.

• Who are the target users for CoCounsel?

  CoCounsel is built for legal professionals, tax professionals, audit professionals, and accounting professionals who require accuracy, nuance, and judgment in their high-stakes work.

• What are the security measures for sensitive work?

  CoCounsel is protected by enterprise-grade privacy and security safeguards, including rigorous privacy and security controls, TLS 1.2 encryption in transit, AES-256 encryption at rest, and zero-retention API calls with AI providers.