The Pros and Cons of Using AI for Multinational Cases | Chapter 4
A Hybrid Approach: Managing Risk with Human Intervention
AI and GenAI present massive opportunities in litigation, as well as several new challenges.
Understanding where and how different AI technologies can help or hinder eDiscovery is essential.
Machine translation has become more common in e-discovery in recent years because of its efficiency.
GenAI has the potential to be a very powerful e-discovery translation tool for cross-border litigation.
Many are turning to a hybrid approach that combines AI and human intervention to improve the efficiency of e-discovery.
AI is here to stay. However, legal teams must still include sector-specific linguists to handle nuanced translations.
Many are turning to a hybrid approach that combines AI and human intervention to improve efficiency while navigating the risks of AI-based translation during e-discovery.
Case Study: Meeting E-Discovery Deadlines with Translation Innovation
That’s exactly what Welocalize did for a Global Am Law 50 firm, which needed to translate 7.5 million words from German to English in less than two months to respond to a second FTC request.
The firm gave Welocalize its e-discovery content, which was processed through MT engines, LLMs, and natural language processing (NLP) technology, then checked by subject matter experts to rapidly automate English content generation from German. Welocalize’s tech-enabled and human-enabled e-discovery solution minimized review errors, increased delivery speed, and decreased discovery costs for the firm.
“Welocalize’s customer service throughout this matter was second to none. Not only did they deliver great value to our firm and e-discovery provider by making the translation process as simple as possible, but the quality of the work product was excellent, and benefited our client’s position by enabling an efficient review by the FTC.”
Associate, Am Law 50, International Law Firm
While AI solutions can eliminate a great deal of risk during e-discovery translation, humans still need to monitor them closely.
Legal teams need to implement quality control measures in which human reviewers check for errors and adherence to industry-specific rules and cultural norms to verify the accuracy of AI-generated translations.
Until AI accuracy improves, it might even be best to reserve AI for efficiency tasks like initial translation drafts and other repetitive, data-intensive tasks where there’s less on the line.
In this vein, Welocalize offers three general strategies for e-discovery translation––all paired with human review conducted by subject matter experts. Those include:
- An out-of-the-box neural machine translation. This option is generally for those who need a general translation that does not need to be certified.
- A customized neural machine translation. This option is generally useful when translations are needed where the language is more technical or culturally nuanced.
- A customized LLM trained specifically on domain-specific data, ensuring translations that are highly accurate and relevant to the industry or subject matter.
- A generative AI engine based on a bespoke LLM.