It is not simply a marker of adoption or growth. It is a sign of trust.��One million professionals chose to trust �������� in��transforming��how they work.��They chose to��test and learn��to rely on something new, and to integrate AI into moments that matter. Each one��represents��a small but meaningful act of transformation: a lawyer who found an hour back in their day, a tax professional who turned research into insight, a compliance officer who had the right information at exactly the right moment. Those moments are why we build.

But if��I’m being��honest, one million is not enough.��We are winning the professional AI market for legal, tax, and compliance—but we are not dominating it.��Not yet. And the gap between those two things is what propels us forward.

Transformation Is a��Journey,��not��a Headline

If you have ever tried to change something fundamental—a process, a company, or a mindset—you know transformation is rarely clean or linear. It is messy. It is slow. It is full of hard conversations and rainy days when true north is hard to find and progress feels elusive.

At ��������, we have experienced all of that. What we have learned is that progress compounds when you keep showing up, block out the noise, and stay anchored to the mission.

Over the past two years, we have been transforming from a historically content-driven company into an AI-powered technology company. That shift takes more than shipping features or adopting new tools. It requires unlearning deeply ingrained habits, questioning long-held assumptions, and building the courage to change how decisions get made.

My role as CTO is to see where change needs to happen, incite it, and steer it—while also��setting��the guardrails and tripwires that keep us from going off the rails. Our��teams’��role is to push against those boundaries and show us when they need to move. That tension between rigor and exploration is where innovation happens.

Creating Sparks of Change��

We see that tension��manifests��most clearly in how change happens. Transformation does not come from large committees or perfect plans. It comes from small, empowered teams making focused progress.

Across ��������, those teams have been the catalysts of change. They moved quickly, tested boldly, learned fast, and shared what worked and what did not. Their work is often unglamorous and invisible, but it is the reason this transformation is real.

They are also the reason CoCounsel exists. They turned agentic AI from a bold idea into something��nearly a��million professionals now use in their daily workflows.

Building Trust from the Inside Out

One of the most important lessons we learned early is that trust cannot be layered on after the fact.��It has to be engineered into the system.

Two years ago, we launched AI Assisted Research—the first generative AI feature in Westlaw. We had a vision of what good looked like, but the reality taught us that defining ‘good’ in generative AI is an iterative process, not a one-time decision.

What felt strong in our research loops needed refinement when put to the test with real human feedback. Legal professionals expected both the precision they relied on and the fluency they were beginning to experience elsewhere. Each round of feedback sharpened our understanding. Each deployment taught us something new about where the bar needed to be.

Those months were challenging, but they were also formative. The conversations with customers and with each other—the honest ones about what was working and what��wasn’t—made our AI more reliable and��reshaped��how we think about accountability in AI systems. We learned how to build solutions with high trust. And in building trust,��slow��became fast.

But over time, this focus on trust created trade-offs we��hadn’t��fully��anticipated. Every verification layer we added, every human review checkpoint, every conservative threshold—they made our AI trustworthy. But they also made us less��versatile,��less ambitious. More precise, but less fluid. More reliable, but less delightful.

We��optimized��for never being wrong. Our users wanted us to also��optimize��for being genuinely helpful.

From Vendor to Partner

Understanding that gap changed how we think about our relationship with customers. We do not want to be another vendor with a product. The world does not need more vendors.

What professionals want, and deserve, is a partner. A partner who listens, adapts, and is honest when something does not work. A partner who understands that trust is earned slowly and lost quickly.

This next phase of our transformation is about moving from transactional relationships to true partnerships. It is about building tools with our customers, not just for them, and meeting them where they are in their own transformation journeys.

Looking Ahead

One million users��proves��we��are��trusted. What it��doesn’t��prove yet is that��we’ve��built the AI professionals genuinely want to use—not just the one they know��won’t��fail them.

That’s��what comes next.��We’re��keeping the trust��we’ve��earned while closing the gap on experience. Being both precise and ambitious. Both reliable and delightful.

This is harder than what��we’ve��done so far. It means moving faster without cutting corners. Being bolder without being reckless. Matching the pace of consumer AI without abandoning professional standards.

But��here’s��what I know: the teams who evolved AI Assisted Research into Westlaw Deep Research—the industry’s most advanced legal research system—and who built CoCounsel into something a million professionals rely on—they’re��not done.��We’re��not done.

Learn more about CoCounsel

But professional work��doesn’t��work that way.

When a lawyer needs to draft a court-ready brief, when a tax professional needs to navigate multi-jurisdictional compliance, when an auditor needs to assess risk across thousands of transactions—they��don’t��need the cleverest chatbot. They need AI that understands their work, their standards, and their accountability.

That’s��what��we’ve��built. And the market is responding.

One million professionals have chosen CoCounsel. Not for pilots. Not for experiments. As core infrastructure for how they work. We serve leading enterprises across legal, risk, compliance, tax, accounting, audit and global trade in 107 countries and territories.

While competitors are��showcasing��demos,��we’re��delivering deployments. While startups are raising capital,��we’re��generating revenue. While others are figuring out trust,��we’ve��already earned it.

The Four Pillars of Professional AI—And Why We Lead

�������� has been building technology and using AI for decades. But what��we’ve��done with generative AI over the past two years puts us in a category of our own.

Professional-grade AI��requires��four essential components working together. �������� has all four��at scale:

We have��the technology. We work with every leading AI lab—Anthropic, OpenAI, Microsoft, AWS, Google—and��we’re��developing our own AI model built specifically for professional work.��We’re��not dependent on a single vendor or locked into one approach. We can use the best technology for each specific workflow.

We have the content.��Decades of curated, authoritative professional content that��can’t��be replicated. Not web-scraped data—the actual��sources��professionals stake their reputations on.

We have��the��expertise. 4,500+ domain experts who understand what “good” looks like in legal, tax, and compliance work. They define quality standards,��validate��outputs, and ensure our AI meets professional requirements.

We have��the tools. CoCounsel integrates directly into the professional workflows and platforms our customers already use—from Westlaw��and��Checkpoint��to Microsoft 365.��Our AI doesn’t sit outside��the��work; it becomes part of��the��work.

These four components working together let us build AI capabilities that others simply��can’t.

Westlaw Deep Research��on CoCounsel Legal��can analyze thousands of documents and synthesize complex legal findings because we combine frontier AI models with our authoritative content library and domain��expertise��to ensure accuracy. Ready to Review��on CoCounsel Tax and Audit��can prepare complete 1040 tax returns—not suggestions or drafts, but finished, filed returns that meet IRS standards—because our tools integrate directly into professional workflows with the quality standards our experts define.

Having one or two of these components means you can build demos. Having all four means you can build products professionals trust with their reputations.

We Have Advantages That Accelerate Our Lead

Having��the four��essential pillars is necessary. But two��additional��advantages strengthen our position:

We have��the��scale. One million professionals using our AI in production��teaches��us things competitors��can’t��learn from pilots. Every edge case becomes common. Every rare failure happens daily. That feedback loop makes our AI��better,��faster.

We have the capital. �������� invests more than $200 million annually in productized AI and��has��11 billion dollars��in capital capacity through 2028 to fund continued innovation and selective acquisitions.

But having all four pillars plus scale and capital only matters if professionals��actually trust��your AI with their work. And trust at this level—where reputations, client relationships, and regulatory compliance are on the line—requires a different approach than consumer AI.

Trust��and Capability:��Our Competitive Advantage

Here’s��what we��learned��building AI at scale: in professional work, you need both trust and breakthrough capability. One without the other��isn’t��enough.

Consumer AI��optimizes for��impressive demos. Professional AI must deliver results you can stake your career on,��while��actually transforming��how work gets done.

Two years ago, we launched AI Assisted Research,��the first generative AI feature in Westlaw. We had a vision of what “good” looked like, but��deploying to��real legal��professionals taught us that defining quality in AI is iterative, not one-time.

Those early months were challenging. Every conversation with customers, every piece of feedback, every deployment taught us something new about where the bar needed to be. We learned that professionals expect both the precision��they’ve��always relied on from �������� and the��transformational capability��they’re��experiencing with generative��AI.

We built for both. And we used what we learned to build something even better.

Every piece of feedback from AI Assisted Research informed how we developed Westlaw Deep Research—now the world’s leading AI legal research capability. Deep Research��doesn’t��just answer legal questions; it analyzes thousands of documents, synthesizes complex findings across��jurisdictions, and delivers court-ready analysis with the citations and reasoning professionals require.��It’s��what happens when you combine frontier AI technology with authoritative content, domain��expertise, and real-world learning from a million professionals.

That same approach drives everything we build. CoCounsel delivers capabilities no one else can match��–��the most advanced legal research system in the world, and the first AI that can prepare a complete 1040 tax return. Not suggestions or��drafts, but��finished work that meets professional standards.��

These��aren’t��incremental improvements.��They’re��capabilities that fundamentally change��what’s��possible in professional work.

And we deliver them with the trust professionals require:��

• ��Accuracy you can stake your reputation on��– because we verify outputs against authoritative sources��
• Transparency you can explain to clients and regulators��– because we show our reasoning and cite our sources��

• Security that guarantees your data stays yours��– because we understand professional confidentiality��isn’t��negotiable��

• Integration with professional workflows��– because AI that sits outside your tools��doesn’t��transform your work��

This is what professional-grade AI means.��Breakthrough capability with professional trust.��And this is why one million professionals chose ��������.

We’re��Accelerating—And Defining the Future

One million users��proves��we’re��the leader in professional AI. But leadership��isn’t��a milestone—it’s��a commitment to staying ahead.

We’re��advancing��development��of our vertically specialized language model designed specifically for legal, tax, and compliance work.��We’re��expanding��CoCounsel’s��global footprint.��We’re��releasing new workflow-specific capabilities throughout 2026 that will further separate us from competitors.

Because��here’s��what we know: the real AI race��isn’t��about who builds the smartest general-purpose model.��It’s��about who can deliver transformational ROI in high-stakes professional environments where trust is non-negotiable.

�������� is��leading��that race. We have��the technology,��expertise, content, tools, scale, and capital. We have one million professionals��who’ve��chosen us as their AI partner. And��we’re��building the best professional AI in the world.

This��isn’t��the beginning of our AI journey—we’ve��been on it for decades. But it is the moment when the market recognizes what��we’ve��become: a leading AI technology company��that’s��defining what professional-grade AI means.

One million professionals are already experiencing that future. And��we’re��just getting started.

Learn more about CoCounsel.����

Tax, Audit and Accounting has always been a profession built on rigor and responsibility. But the reality of running a modern practice is complexity keeps rising, timelines keep tightening, and client expectations keep evolving. Firms are being asked to do more, faster—without compromising quality.

Why this matters for firms

This recognition is not just about a new solution being introduced—it’s about what firms are prioritizing right now. Across the industry, leaders are focused on creating repeatable capacity—not just surviving busy season. They’re prioritizing workflows that are reliable year-round. That means reducing friction in the return process, improving consistency across teams, and ensuring professionals have the time to apply judgment where it matters most.

That’s the lens I bring to . It’s designed to support the parts of the tax return process that can slow teams down, so professionals can stay focused on the work of humans: review, judgment, accountability, and advising clients with confidence.

Restoring time for professional judgment

There’s a lot of discussion about automation and the future of work. What I see in firms today is more practical: talented professionals spending too much time on repetitive steps, and not enough time on review, coaching, and client conversations.

The opportunity here is to shift time back to the professional—so firms can:

• Strengthen quality and consistency
• Improve responsiveness to clients
• And make the work more sustainable for teams

Honorable Mention: Ready to Advise

Accounting Today also gave an Honorable Mention to , which supports tax planning and advisory services.

That matters because once firms create more capacity, the next question is how to use it. Many firms are looking to grow advisory in a way that’s scalable and consistent – grounded in strong workflows and clear client outcomes.

Recognition like this is meaningful for us at ��������. But it’s even more meaningful because it reflects progress our customers can feel. When firms can rely on their technology to create more capacity and consistency in the work, they can serve clients with greater confidence. They can support their teams through peak demand, and make the practice more sustainable. That’s the kind of win we’re focused on: one that strengthens firms and the professionals who power them.

Elizabeth Beastrom is President, Tax, Audit & Accounting Professionals at ��������

��

Ready to Review is our new cloud-based, agentic AI tax workflow solution built on CoCounsel. For tax year 2025, it modernizes 1040 tax return preparation. It’s designed to take on the heavy, repetitive work of gathering client documents and preparing returns—so tax professionals can focus on what drew many of us to this profession in the first place: problem solving, critical thinking, and delivering great counsel to clients.��

For too long, firms have been stuck in a pattern that everyone recognizes as unsustainable—ever more complex returns, tighter deadlines, and mounting pressure on teams already stretched thin. Busy season has become synonymous with burnout and staffing strain. AI alone won’t fix that. But AI put to work in the right way—through agentic AI deeply embedded into tax workflows—can fundamentally change the equation.��

A Better Way Through Busy Season

The goal is not to replace professional judgment, but to clear away the manual, time-consuming tasks that prevent professionals from using that judgment to its fullest. Ready to Review gives firms a single, cloud-based, scalable platform that automates the gather and tax prep stages of the��workflow—helping firms manage more individual returns with existing staff while��maintaining��quality and control.��

We’re��already seeing the impact. Indiana-based CLH CPAs & Consultants��participated��in our early adopter program and saw��the potential for��transformative time savings. As Bob Lange, Partner at CLH, told us:��“Reducing return preparation time by��approximately an��hour on each simple 1040 is significant in terms of efficiency gains. For firms like ours, these time savings will be a game changer.”��

I’ve��said before that I expect firms will��ultimately pair��each CPA with at least one virtual agent. is a tangible step in that direction. It brings that 1:1 vision closer to reality by embedding Gather and Tax Preparer AI agents directly into the 1040 workflow in a way��that’s��responsible, auditable, and grounded in trusted �������� tax content and compliance��expertise.��

The Next Step for 1040 Prep

As solutions like Ready to Review become part of the day-to-day fabric of tax work,��we’ll��see fewer 80-hour��weeks��and more time spent on the nuanced, client-focused work that truly differentiates firms.��We’ll��make room for new talent who are excited about a career that leans into analysis and advisory rather than pure grind.����

Ready to Review is now generally available in the United States for 1040 use cases, and its launch marks an important milestone in our broader journey with agentic AI on the�� platform.��It’s��one more way��we’re��helping firms modernize in a way that is practical, grounded, and built for the realities of tax season.��

Elizabeth Beastrom is President of Tax, Audit & Accounting Professionals at ��������

At ��������, our starting point is, and will remain, methodology. For decades, firms have relied on �������� PPC methodology as the gold standard for audit quality, documentation, and compliance. Our vision for AI in auditing builds on that foundation. We’re not asking firms to change how they practice. We’re focused on making PPC the most AI‑automated audit methodology in the market—through our own products and through deep partnerships with innovators our customers already trust.

That’s the idea behind our recent partnerships with , , , , , and . Together, we’re embedding PPC into AI‑driven tools across We’re supporting AI-powered automation with , so firms can automate more work while staying grounded in the trusted methodology they already rely on.

Trullion: Methodology‑aware automation for financial statement review

Financial statement review is one of the most judgment‑intensive parts of the audit—but many of the underlying procedures are repeatable. Our integration with Trullion brings AI‑native automation to financial statement review and testing, with full traceability back to PPC methodology and the relevant guidance at every step.

Artie Minson, CEO at Trullion, describes the shift: “This partnership signals a new era for audit automation and lays the foundation for trusted and truly agentic workflows. Our vertical AI solution is built for auditors by auditors, ensuring our outputs are within the framework of professional standards. This integration creates methodology-aware automation. Auditors can now focus their time on applying judgment to fully evidenced, agentic outputs, rather than searching for them, delivering audits with unmatched efficiency, accuracy, and quality.”

For us, “methodology‑aware” is key: automation is valuable only when it operates within the same professional framework firms already use to define, document, and support their work.

Audit Sight: Substantive analytics that reduce testing

Substantive testing is another area where firms feel the strain. Even when technology is available, many teams still default to large samples and manual procedures.

As T.C. Whittaker, Co‑Founder and CEO of Audit Sight, puts it: “Audit firms are seeking smarter ways to expand capacity and elevate quality without adding headcount. Bringing automated testing together with ��������’ PPC methodology — and enabling it through Guided Assurance — is the ultimate unlock for auditors. It transforms the audit plan itself, making it intelligent and dynamic by tailoring procedures, eliminating unnecessary tests, and reducing sample sizes based on automated evidence and client-specific risk. This partnership represents a shared vision to redefine how assurance is delivered in the modern era.”

Crunchafi: Automating lease procedures inside PPC

Lease accounting has become a complex, time‑consuming area for many firms. Too often, teams spend hours on calculations and reconciliations instead of higher‑value work.

By integrating Crunchafi into Guided Assurance, we bring seamless lease accounting automation directly into PPC‑based workflows, eliminating manual lease calculations and providing audit-ready journal entries, amortization schedules and footnote disclosures while preserving firms’ established methodology.

Mike Cooke, CRO at��Crunchafi, explains: “Audit teams want efficiency without sacrificing quality. By aligning��Crunchafi’s automation with the PPC Methodology, we’re giving firms a clearer, more reliable way to handle lease accounting from the start of the engagement to the final deliverable.”

This is the pattern we’re aiming for: automation that plugs into how firms already work, rather than asking them to start from scratch.

Fieldguide: Empowering Firms with Flexible Paths to Automate PPC Methodology

Many firms also want a more connected environment where methodology, evidence, and automation all live together. Our goal is to meet firms where they are – and give them options.

That is why we’ve partnered with Fieldguide to embed Guided Assurance—which delivers PPC methodology—directly into Fieldguide’s professional‑grade agentic AI platform. This creates a unified experience where trusted PPC content and intelligent automation collaborate to execute engagements efficiently and consistently.

Whether firms choose to automate audits with �������� or Fieldguide, they can be confident they’re using the most trusted and automated methodology in the profession. This flexibility reflects our commitment to innovation and the unique needs of our customers.

Jin Chang, Co-Founder and CEO of Fieldguide says: “Firms are under pressure to do more with less. They need trusted methodology and AI agents that work the way they do. By embedding �������� PPC methodology into our platform, we’re helping firms deliver higher quality work with more consistency and less effort. This partnership reflects a shared commitment to the future of the profession.”

Validis: Data as the foundation for AI‑driven auditing

AI is only as good as the data behind it. For many firms, getting clean, audit‑ready data from clients is one of the toughest operational challenges.

Through , our work with Validis focuses on solving that. Validis powers secure, on‑demand ingestion of client trial balance, general ledger, and subledger data directly into Audit Intelligence. From there, we use AI and machine learning to focus testing on high‑risk areas, segment populations by risk, and reduce the number of items to be tested, with anomaly detection automatically surfacing unusual items and generating the required documentation.

As Jeff Gramlich, Managing Director at Validis, explains: “We’re excited to collaborate with ��������, a true market leader and innovator, to deliver audit-ready data through our cutting-edge ingestion capabilities. This partnership provides auditors with the data breadth and granularity crucial for effective AI-driven auditing. By integrating our technology into the Audit Intelligence suite, we’re empowering auditors to conduct data-driven audits with enhanced efficiency and risk analysis, ultimately transforming the process to benefit both auditors and their clients.”

Valid8 Financial: Turning evidence gathering into an automated workflow

Finally, there’s the everyday work of matching samples to evidence and documenting that work in a way that stands up to inspection and peer review. This is some of the most manual and time‑consuming work in an audit.

, developed with Valid8 Financial, automates the matching and documentation of samples to supporting evidence, dynamically tracing accounting transactions to banking activity to confirm occurrence. It brings technology traditionally used in advisory, forensic, and financial crime work into an integrated audit workflow.

Brett Suchor, CEO of Valid8 Financial, says: “We built our technology to solve real problems auditors face every day – reducing the manual, time-consuming work of matching samples to evidence. Through our collaboration with ��������, we’re delivering a faster, more reliable testing experience to audit professionals across the industry.”

The future of Audit

In the next 3-5 years, we’re going to see big changes to the audit profession. Audit is moving decisively toward an automated, data-driven future. Using the right tools to increase efficiency and quality so teams can focus on higher risk areas and deliver better outcomes for clients is paramount.

In today’s environment, firms are being asked to do more with less, navigating tighter deadlines, increasing complexity, and growing client expectations. At ��������, we are bringing auditors advanced audit technologies, with our��newest audit solutions��increasing efficiency and accuracy.

We’ll keep investing in our own AI capabilities and in this partner ecosystem so firms can modernize at their own pace, on their own terms—without walking away from the methodology that has served them well for decades.

This post was authored by Dave Wyle, General Manager of Audit at ��������.

At TechCrunch Disrupt 2025, the buzzword wasn’t “AI.” It was scale. How do you take technology powerful enough to transform billion-dollar industries and make it trustworthy enough to run them?��

That’s the challenge �������� has been solving in real time.��

At the Women of Disrupt Breakfast: From Vision to Velocity, Women Driving AI Innovation at Scale, Laura Safdie (Head of Legal Innovation, ��������, and former co-founder of Casetext) and Kirat Sekhon (Head of Engineering, ��������) joined Martine Paris, Forbes and BBC AI reporter, for a conversation on building agentic AI that doesn’t just assist professionals, but collaborates with them.��

From Startup Grit to Global Infrastructure��

Laura Safdie knows what it means to build from scratch. Before joining ��������, she co-founded Casetext, the legal AI startup that created CoCounsel, the world’s first GenAI legal assistant.��

“When GPT-4 launched, we knew the ground had shifted,” Laura said. “The world, and our profession, would never be the same.”����

Within a week, Casetext rewrote its roadmap and shipped a working product that redefined legal work. Months later, �������� acquired Casetext, turning that same startup innovation into the foundation for a professional-grade AI ecosystem now used across legal, tax, and corporate domains.��

Today, CoCounsel powers workflows for hundreds of thousands of professionals worldwide, combining the speed of machine learning with the rigor of human judgment.��

The Next Evolution: AI as the Junior Professional��

Forget chatbots. The next era of AI is here, and it looks a lot like your most capable new hire.��

In law, that means systems that can draft, review, and analyze complex documents with context and accuracy. In tax and accounting, it means interpreting new regulations, scanning data sets, and preparing the groundwork for filings at lightning speed.��

“The human is still the strategist,” Kirat explained. “But the AI is that relentless team member who never tires, never loses focus, and helps you get to the insight faster.”��

This isn’t automation. It’s augmentation. It’s about freeing people to do the creative, analytical, and human work that truly moves the needle.��

Why Trust Is the Killer Feature��

In high-stakes industries, accuracy isn’t optional. “Close enough” doesn’t cut it.��

That’s why trust has become the new measure of technical excellence. �������� builds AI that shows its work, with authoritative citations, verifiable sources, and a full digital audit trail.��

Whether it’s a contract analysis or a tax interpretation, professionals can trace every step. Transparency isn’t an add-on; it’s the architecture.��

In a world where hallucinations can tank credibility, professional-grade AI earns trust one verified line at a time.��

Building Fast Without Breaking What Matters��

Innovation moves at a blistering pace. Models update weekly; frameworks shift overnight, and what’s state-of-the-art today can feel dated tomorrow.��

That’s why �������� has engineered adaptive architecture. A flexible layer that lets teams integrate new large language models, swap them out, and test emerging capabilities with precision and control.��

Every model passes through a rigorous evaluation framework that measures accuracy, speed, and relevance. Engineers even collaborate directly with model developers, shaping future releases with real-world performance data.��

But speed isn’t just about shipping code. It’s about changing how people think and work. “This is the most energizing moment in many people’s careers,” Laura said. “But it takes a mindset shift. Change management is as important as the technology itself.”��

Resilience as the Invisible Superpower��

When cloud providers falter or networks crash, professionals still expect their tools to perform. That’s why resilience is built into every layer of the �������� AI stack.��

The company’s systems are multi-cloud and multi-model by default. If one model slows, another takes over. If a provider fails, others stay live. It’s a design built for continuity, transparency, and trust in motion.��

Reliability is no longer a technical metric. It’s a brand promise.��

The Human Edge����

�������� is hiring AI engineers and data scientists who want to build the future of agentic systems. AI that collaborates, learns, and adapts alongside professionals.��

The goal isn’t to replace human expertise. It’s to elevate it.��

The future of professional work will belong to teams that combine computational intelligence with human judgment, creativity, and integrity. AI isn’t just transforming how we work; it’s transforming how we lead.��

The Takeaway��

The companies that win in this new era won’t just move fast. They’ll build right.��

The next wave of AI innovation will be defined by systems that scale, teams that adapt, and leaders who build with trust and purpose.��

As Laura and Kirat reminded the audience at TechCrunch Disrupt, the goal isn’t just smarter technology. It’s a smarter, more human future for professionals everywhere.��

Why legal input is crucial

The law is a complex universe – each case requiring careful and thorough research, each document requiring specific language to ensure its validity and compliance. When lawyers leverage legal technology to find answers, assist with drafting, or carry out multi-step workflows, they need to feel confident that they’re not missing something. Many of the tasks we expect legal AI to perform involve a nuanced understanding of the law and its application, which is often anything but black-and-white. Without meticulous testing and grading by legal experts, the door is left open for costly mistakes. Take the for example. When testing various legal AI platforms against a human lawyer, 3 out of 4 platforms could not identify and extract contract language relating to a specified clause. The one that did? �������� CoCounsel.

What this looks like in practice

�������� is deeply committed to using legal and subject matter experts when testing and grading the output of our legal AI.

For our developers working on AI for legal research, that means partnering with our licensed Westlaw attorney editors. Our editors help identify data that should be referenced for each skill, create standards by which to test the AI, conduct the AI testing, and evaluate and grade the output. Our attorney editors have conducted hundreds of evaluation sessions and graded thousands of responses to ensure that our AI-enhanced tools like CoCounsel meet accuracy standards, adhere to source documents, and apply logical reasoning that accounts for nuances in the law.

Similarly, our Practical Law attorney editors are integral in our development and improvement of agentic workflow capabilities in CoCounsel and beyond. Not only do our subject-matter experts create gold-data tests for our new agentic capabilities but also improve how we conduct human grading and output evaluation for autonomously performed complex, multi-step legal tasks.

CoCounsel’s outputs must meet complex criteria like factual accuracy and logical consistency, which aren’t easily judged with simple true-or-false tests. Evaluating legal content is also often subjective – some users prefer detailed summaries, others concise ones—making automated assessments challenging. Enter our Trust Team. The Trust Team is a group of experienced legal professionals with backgrounds ranging from in-house counsel to law firms of every size. They create tests that represent actual work attorneys need to complete, set up the gold-standard response, and then run these tests against CoCounsel’s skills for automated evaluation. Using this process, CoCounsel’s skills have undergone over 1,000,000 tests, and any output that does not meet the attorneys’ standards is reviewed manually to ensure reliability.

Bringing it all together

The recent development and release of the CoCounsel workflow Draft a Discovery Request highlights the importance of collaboration between tech and legal expertise. It’s not enough to build an AI tool with the hope it might be useful to attorneys. You need actual lawyers to tell you what a specific workflow looks like now, what their pain points are, and how AI can alleviate those challenges.

When creating Draft a Discovery Request, the CoCounsel team relied on three separate teams of legal subject matter experts to guide the build: seven litigators from the Trust Team to create consistent automated and manual testing during development, an AI editorial team to provide process and grading input, and 13 Practical Law editors with litigation experience ranging from labor and employment to intellectual property law to review formatting, style, and the substantive output. During the testing and review process, the Practical Law editors noted that the skill output omitted several key definitions and neglected to include requests relating to interstate commerce issues and additional parties connected to the suit. Any one of these errors could result in confusion or an objection from opposing counsel, but careful editorial review allowed our CoCounsel team to adjust the output requirements and account for the necessary legal considerations and best practices. Without the review of the 20 legal subject matter experts working on this skill, these issues would not have been flagged because they were not technical errors, but substantive or procedural ones. Legal expertise and guidance were critical to verifying the accuracy of the workflow and ensuring this is AI lawyers will find useful and trustworthy.

AI that thinks like a lawyer

Having attorneys and legal subject matter experts provide guidance, develop test criteria, and evaluate legal AI is critical to providing a truly valuable tool for legal professionals. You can’t have AI that thinks like a lawyer if no lawyers are involved in the process. That’s why CoCounsel Legal is the AI lawyers swear by.

��

Introduction����

At ��������, we are redefining what it means to deliver professional-grade AI for the legal industry. More than 20,000 law firms, corporations, nonprofits, and government agencies worldwide rely on CoCounsel, our GenAI assistant, which transforms how legal professionals work by automating complex document review, contract analysis, drafting, and other time-intensive tasks with unprecedented speed and accuracy. That trust is earned through a comprehensive evaluation methodology that encompasses dataset rotation, automated testing, expert assessment, continuous monitoring, and strategic partnerships. This post focuses on one critical component of our broader testing framework: how our teams combine attorney expertise with large-scale automated testing through Scorecard, a proprietary evaluation platform originally developed by the engineers behind Waymo’s self-driving car testing infrastructure. While Scorecard represents just one pillar of our multi-layered approach, it exemplifies our commitment to proactive system optimization and continuous improvement. ��

Testing and Benchmarking with Scorecard��

Our teams of attorney subject matter experts (SMEs), machine learning experts, and engineers rely on a robust array of testing tools and methodologies, including human legal expertise, specialized testing software, expert prompt engineering, and continuous monitoring of test results. Rather than waiting for performance issues to emerge, we proactively identify and address potential challenges through systematic testing and optimization. When issues arise that may affect CoCounsel’s performance, these teams are equipped to mobilize a collaborative, rapid response effort, locating and remedying performance issues before they affect our customers.����

A key tool is Scorecard, a specialized application that quantitatively evaluates CoCounsel responses against ideal responses created by our attorney SMEs. is the evaluation infrastructure for AI agents in legaltech, fintech, and compliance, and enables us to supplement our manual testing with large-scale, automated testing against our internal benchmarks. Built by the team behind Waymo’s self-driving evaluation infrastructure, Scorecard runs millions of agent simulations to help teams evaluate, optimize, and ship reliable AI agents faster.��

Performance issues typically arise from two distinct factors:
(1) the quality of user inputs, such as user prompts or queries, and documents; and,
(2) system limitations.

We address the first factor by providing customers with high-quality training, support, and tools—including CoCounsel-created prompts, guided expert workflows, and agentic systems. In contrast, addressing the second factor requires recalibrating the system itself.��

Each CoCounsel skill is a precisely engineered legal tool, tailored on the backend to perform a specific legal task. Because we calibrate each skill to reliably extract information by leveraging the unique strengths of its underlying AI model, migrating a skill from one model to another often introduces performance issues that require recalibration. Such migrations may occur, for example, when a third party releases a new AI model with enhanced capabilities. To safeguard our customers, we conduct all migration and recalibration work within testing and staging environments before deploying any changes.��

Case Study: AI Model Migration of Review Documents Skill��

Large-Scale Testing Using Realistic Scenarios & Manual and Automated Review��

Jessica, an attorney SME on the CoCounsel AI Reliability Team—also known as the Trust Team—oversees the evaluation of CoCounsel’s Review Documents skill. In just minutes, the Review Documents skill can closely review and analyze large troves of legal information that would ordinarily require an attorney to spend hours or even days of manual review.����

Jessica proactively monitors the upcoming migration of the Review Documents skill to a new AI model. This migration promises significant improvements in CoCounsel’s speed and accuracy. Working in a CoCounsel testing environment, Jessica manually reviews and evaluates the skill’s responses on the new model using a carefully curated “testset” of sample “testcases” that reflect real-world legal practice scenarios. Jessica checks CoCounsel’s response to each testcase user query against an ideal or “gold-standard” response that she has personally crafted using knowledge and expertise gained from years of experience as a real-world attorney.����

Because each testset can contain several hundred testcases or more, reviewing each result would ordinarily be prohibitively time-consuming. However, Scorecard enables Jessica to supplement and scale the impact of her manual review by providing an extra layer of automated review.����

Scorecard works by evaluating each response produced by CoCounsel and the AI model against the corresponding ideal response, then assigning the testcase a passing or failing numerical score using several criteria, such as the model’s ability to recall information, its precision, and its accuracy.����

Reviewing the Scorecard results enables Jessica to compare the full testset’s scores on both models for the Review Documents skill. This means she can evaluate CoCounsel’s performance at scale much more efficiently.

Fig 1: Attorney SME manual review workflow.

Fig 2: Scorecard automated review workflow.

Reviewing the Scorecard data, Jessica quickly observes that on the new model, Scorecard consistently assigns failing scores to a specific testcase, assigning it a 1 out of 5 on all metrics. She identifies underperformance in other testcases, too; however, the other testcases still yield higher scores than the problem testcase. Recognizing the stakes are high, Jessica immediately begins troubleshooting the performance issue.

Troubleshooting

Jessica and her team of SMEs begin to troubleshoot by homing in on the problem testcase that Scorecard identified.

The testcase user query asks:

What medications is the patient currently taking? Please be specific with prescription names and dosages.

Analyzing CoCounsel’s outputs for the testcase, Jessica determines that on the new model, the Review Documents skill is failing to identify all medications for the patient consistently, causing a clear discrepancy with the ideal response. The new model occasionally includes all the relevant medications, but such inconsistent behavior does not meet the required standard.

[Click image to expand] Fig. 3: Scorecard screenshots of the AI model’s failing answer. As can be seen in the expanded “model response” window above, the model was including medications that were no longer currently active and was failing to identify the only two current, active medications (Aspirin 81MG EC TAB and Aspirin 325MG EC TAB).

By digging deeper and examining the problem testcase response as well as some of the other, underperforming testcase responses, Jessica pinpoints the core issue as being the AI model’s ability to provide a sufficiently comprehensive level of detail. Since the model sometimes does output a complete response, Jessica observes, as a secondary concern, that the AI model struggles to produce consistent results.

Iterative Resolution & Continuous Improvement

Having identified the core issues, Jessica brings the issue to the CoCounsel engineering team for resolution. She describes the parameters of an ideal response and how the new model’s response fails to meet target metrics. This gives the engineers concrete goals, which they can use to modify the backend AI prompts. After each prompt change, Jessica evaluates a portion of the test set which is continuously updated, complemented by independent attorney reviews. Jessica and the engineering team continuously execute multiple rounds of prompt changes and use Scorecard to evaluate the results until the issue has completely resolved, and the new model is performing as expected. Scorecard now assigns the problem testcase a 4 out of 5 on all metrics, a good score—it reflects that the model has produced a valid response that captures all relevant substantive data points contained in the ideal response but may differ in more subtle ways, such as writing style or level of additional detail. Resolving this core issue ensures the secondary issue of inconsistent performance has been resolved as well. Jessica further conducts manual reviews of CoCounsel’s performance on the problem testcase.

These adjustments have cascading positive effects. When the problem testcase begins passing 99-100% of the time, the other testcases that had experienced the same issues (albeit less frequently) begin passing 100% of the time.

[Click image to expand] Fig 4: Scorecard screenshots of the AI model’s passing answer. This was achieved after multiple rounds of testing and prompting changes, which confirmed the engineers were able to pinpoint and fix the issue. As shown in the expanded “model response” window above, the issue was ultimately fixed, and the model began answering this testcase correctly (as well as a few other testcases that had been failing, albeit less frequently, due to the same issue).

Once the model consistently returns results that meet TR’s expectations and are suitable for legal work, Jessica feels secure in the knowledge that the Review Documents skill meets necessary standards and can be released to customers.

Even after the skill is released on the new model, Jessica continues to run various Scorecard tests, multiple times daily to ensure consistency.

Fig 5: Continuous improvement process between attorney SME and engineers.

Observations

CoCounsel’s proactive and continuous iterative improvement process is painstaking but necessary. The problem testcase identified by Jessica using Scorecard provided a useful benchmark for improvement, because it failed more consistently than other testcases. Using a “least common denominator” testcase provided a measuring stick against which we could measure other testcases.

Using Scorecard allowed Jessica to extrapolate improvements from the single problem testcase to all other testcases, dramatically increasing the efficiency and speed with which she could iterate and improve CoCounsel’s performance across the board.

Conclusion

Innovation in AI is never “one and done.” Models evolve, new risks emerge, and customer needs grow more complex. While this post has focused on Scorecard as one essential component of our testing infrastructure, it represents just one element of our comprehensive evaluation methodology. Our broader approach integrates dataset rotation, automated testing at scale, expert assessment from legal professionals like Jessica, continuous monitoring of live performance, and strategic partnerships with leading AI providers.

This multi-layered framework is what sets CoCounsel’s approach apart. By combining deep legal expertise with world-class technology infrastructure, we’re not only raising the standard for AI in professional fields, we’re defining it. Through proactive system optimization and evaluation approaches, CoCounsel continues to deliver the transformative professional-grade legal AI capabilities that tens of thousands of legal professionals depend on.

—-

About the Authors

Tyler Alexander is the Director of AI Reliability at ��������, where he leads a team of attorneys to ensure CoCounsel delivers trustworthy, professional-grade performance. He specializes in large-scale testing and benchmarking of AI systems for legal professionals.

Heather Nodler is a Lead CoCounsel AI Reliability Manager at ��������. With years of experience practicing law, they now apply their expertise to evaluating, calibrating, and continuously improving CoCounsel’s legal AI skills. Heather works closely with product and engineering teams to ensure that every CoCounsel feature meets the high standards required for real-world legal practice.

Redesigned drafting capabilities unify CoCounsel tools, content, expertise, and workflows

Informed by customer feedback, we’ve reimagined the legal drafting experience – making it more intuitive, intelligent, and seamlessly integrated. The drafting capabilities in CoCounsel fuse users’ institutional knowledge with trusted �������� content and AI-powered technology to expedite the legal drafting process. The redesigned homepage puts everything users need right at their fingertips – CoCounsel Chat, skills, and powerful litigation and document analysis tools – all in one clean, intuitive space. Eliminating the need to jump between tabs or hunt for resources, it’s now an even smoother, faster experience that lets users stay focused, work smarter, and get more done with less friction.

Live Draft brings the ability to summarize and modify a document using natural language in Word. Live Draft also delivers contextual awareness of the document and understanding of the content and structure, so every suggestion and edit is tailored to the content. This helps to further reduce time spent producing a final draft, by delivering more accurate, relevant suggested changes.

Append Authorities enables users to combine all cited documents into a single file suitable for court use, reducing the risk of errors and increasing efficiency. Every cited document is linked for verification purposes, and a hyperlinked table of contents is included.

Region settings customizes CoCounsel tools for global legal professionals

The new region settings capability enables users to select their geographic preference from U.S., UK, Australia or Canada. Based on the selected region, region settings will automatically adjust tools and prompts in the CoCounsel Library making the work product more relevant. Users can now automatically tailor their documents using specific regional requirements, including for the UK and Australia, British English spelling variations, legal terminology, grammar, and content formats. Similarly, this will be coming soon for Canadian English. Additionally, CoCounsel Library is now available in the UK, Canada and Australia.

HighQ integrates CoCounsel AI for intuitive, conversational client data access

With��CoCounsel’s Search a Database skill embedded within HighQ, this customer-driven development allows clients the ability to pose queries regarding their data and receive summarized, highly relevant answers. Sourced from pre-approved content within their site, clients can quickly review summaries, generate reports, and make informed decisions without waiting for manual responses.

Legal Tracker adds AI-powered capabilities

Legal Tracker’s new AI features help users manage legal spend more efficiently. The AI-powered PDF-to-LEDES converter and invoice review speed up invoice evaluation and ensure accurate billing. An AI assistant also streamlines reporting and reduces manual data handling.

These transformative features reinforce our commitment to empowering legal professionals with the tools and solutions they need to excel. or to see firsthand how they elevate work to new heights.

To stay abreast of newly added features, monthly releases, and more, please sign up for the .

But what followed was less about innovation and more about disruption. Longstanding models like GPT-4o were pulled without warning. Familiar workflows broke. Performance felt inconsistent. Some users even said the model felt distant and robotic. Within days, OpenAI had rolled back several changes and re-enabled access to older models.

It wasn’t a failure of any one model or company but rather a failure of expectations. And it’s a reminder of a broader truth in the AI industry: even the most advanced systems can introduce friction if change outpaces the ability to adapt to it. As models evolve, so must the frameworks around them, especially in professional environments, where progress only matters if it delivers measurable, reliable benefits for the humans it’s meant to empower.

At ��������, we work with lawyers, tax advisors, and compliance professionals whose work leaves no room for guesswork. For them, consistency is not a preference—it’s a fundamental requirement for them to uphold their professional duty to their clients. That’s why we don’t chase upgrades for their own sake. And we certainly don’t ask our customers to pick which model they want to use. That’s our job. Our customers expect us to deliver a result they can trust, not a menu of models to experiment with. They want confidence, not complexity.

When we evaluate a new LLM, we do it through the lens of real-world use:

• • Can it reason over long documents with accuracy?
  • Can it explain its conclusions with transparent citations?
  • Will it behave consistently inside multi-agent workflows?
  • Does it integrate with how professionals already work?

If the answer is no, we don’t ship it…until we’re confident that we’ve mitigated those concerns appropriately.

One example: earlier this year, our team benchmarked several leading LLMs for long-context performance. The task was to extract and apply insights from large, multi-thousand-word legal documents, a common need in law and compliance. We found significant variance. Some models struggled to maintain context or reference earlier sections accurately. Others returned plausible-sounding answers that fell apart under scrutiny. Rather than push forward with the best-performing model, we paused. We refined how our agents chunk and reason over large documents. We optimized prompts and guardrails. And we only moved forward when the system delivered answers that we’d be willing to stand behind in a courtroom.

This kind of work doesn’t show up in a product demo. But it’s what builds trust.

We also design our products to abstract that complexity away. In CoCounsel Legal and Deep Research, we use multi-agent systems to coordinate model selection, content access, and validation behind the scenes, so the user sees a transparent, explainable result, not a swirling mix of models and prompts.

Recent model rollouts offer an important reminder: in enterprise AI, newer isn’t always better. Progress should be measured not just by technical benchmarks, but by the clarity, consistency, and confidence it delivers to real users. The systems that will define the next chapter aren’t just the most advanced, they’re the ones that work reliably, integrate seamlessly, and build trust from day one.

The reality is, there will be more disruption. We are all moving fast because the potential of AI is enormous and the demand for it is real. But speed does not have to come at the expense of the hard-earned trust of our customers. The more we treat disruption not as a cost of innovation, but as a signal to improve our processes—model governance, human oversight, testing frameworks—the better we will get at delivering AI that is not just powerful, but trustworthy. Over time, the industry will learn. We will see fewer rollbacks, clearer standards, and smarter integration. But that will only happen if we choose to build that way, with intention, transparency, and the end user in mind.

That’s the future we’re building toward. Not hype-proof. Trust-proof.