Blog Post

In 2024, the legal industry reached a tipping point. After reviewing the speed with which generative AI (GenAI) could review large data sets, and seeing that GenAI was better than human reviewers at categorizing documents, law firm partners quickly realized that GenAI would be the future of first-level document review.  Initial skepticism about data security and accuracy gave way to tangible results — faster workflows, cost savings, and higher accuracy than human reviewers could achieve alone.

By 2025, what was once a whisper in the hallways of AmLaw200 firms is becoming an open secret: GenAI is reshaping the document review process, and outsourced review as we know it may soon be a relic of the past. Rather than outsource review to managed review providers, law firms are using platforms that enable automated first-level review, like Syllo, to take large reviews back in house.  Smarter, more targeted review has meant that firms can bill attorney time for factual investigation.  Revenue that once flowed to third-party managed review firms is now flowing back into law firms.

This transformation has been made possible by the unmatched speed and accuracy of Generative AI review. Use of models enables firms to process thousands of documents per hour, allowing legal teams to meet tight deadlines without compromising quality. This capability outpaces anything achievable by human reviewers. Firms like Ballard Spahr LLP have leveraged GenAI to identify document deficiencies within hours, providing critical insights earlier in the litigation timeline. This speed ensures that teams can act decisively and strategically, enabling legal professionals to focus on strategy, not manual tasks. 

The rise of GenAI is shifting the role of document reviewers. For review of incoming productions, there is no longer a need to spend manpower on initial document coding.  Case teams immediately search through categorized documents and start their factual investigation.

For outgoing productions, reviewers will now need to develop a unique skill set that blends legal expertise and technical proficiency. Reviewers must have a deeper understanding of case facts and legal principles to evaluate GenAI’s tagging decisions effectively. These reviewers will now conduct quality control analysis to ensure that tagging results fall within acceptable tolerances.

These tolerances are typically substantially higher than older review methods.  Users of Syllo can expect to identify 95% or more of all relevant documents with the initial review already resulting in a high recall rate.  By contrast, initial recall rates for a human review typically top out at 80%.  While reviewers can increase recall rates using traditional technology assisted review, these technologies require iterative rounds of review to increase the percentage of relevant documents identified.

Conducting quality control of GenAI-tagged documents requires reviewers to identify instances of potentially overinclusive or underinclusive tagging and analyze GenAI explanations to determine the basis for any inaccuracies. Reviewers will then provide feedback to optimize a second round of searches.  Thus, rather than producing the tags themselves, reviewers will work closely with the technology to optimize its performance.

To what extent will law firms use their own personnel to conduct these quality control roles or outsource this role to third-party providers? The answer remains unclear. Some firms may prefer to keep AI-driven workflows under their direct control, leveraging their familiarity with the case. Others may rely on third-party providers to manage the technology, offering specialized expertise in AI oversight. What’s certain is that the landscape of outsourced review will look very different in the years to come.

If you’re eager to make sure that your firm is working with the most advanced AI tools to conduct document review, schedule a demo today and see how Syllo delivers tagged documents efficiently, accurately, and cost effectively.