Legal Hold Management

Caveats to consider when opting In-Place Preservation for Legal Hold

We have heard a lot about In-Place preservation being the most proactive method to ensure you never fail to adhere to legal hold orders. Unlike the traditional legal hold, In Place Legal Hold is predicted to be more efficient. In-Place preservation sounds wonderful as you don’t have to incur the collection and storage costs until necessary. But very few discuss the practical implications of adopting In-Place Legal Hold for the eDiscovery needs.

What is In-place Preservation?

In the context of an in-place legal or litigation hold, organizations implement technical measures to prevent the alteration or deletion of relevant data in its original location instead of physically collecting and moving data to a separate storage location. 

This approach allows organizations to avoid disrupting normal business operations and potentially reduce data collection and storage costs

However, implementing an in-place legal hold has certain caveats and considerations that organizations should be aware of. Here are some common caveats associated with in-place legal holds:

Preservation Risks

While in-place legal holds aim to preserve data in its original location, inherent risks are involved. For example, accidental deletion, intentional tampering, or technical failures such as connectivity issues could still result in losing or altering relevant data. Organizations must carefully assess their technical measures’ effectiveness and reliability to mitigate these risks.

System Complexity

In-place legal holds often require coordinating various technical systems, such as email servers, document management systems, and content collaboration platforms. Managing these systems and ensuring consistent application of legal hold measures across different platforms can be complex and may require specialized expertise.

Metadata Management

Metadata, such as file creation dates, author information, or system logs, play a crucial role in maintaining the integrity and context of data. However, preserving metadata accurately during an in-place legal hold can be challenging. Organizations should ensure their preservation mechanisms account for metadata and prevent its inadvertent modification or loss.

Data Overwrite

In certain circumstances where cloud data is involved, data may be automatically overwritten or modified by normal business operations, system updates, or backup processes. Organizations should carefully evaluate the risks associated with data overwrite and implement measures to prevent the loss of relevant information during the legal hold.

Data Accessibility and Organization

As the volume of data grows within an organization, locating and accessing specific relevant information during the legal hold period can become challenging. It is crucial to have robust data management practices, such as effective data indexing, search capabilities, and clear identification of custodians and their responsibilities.

Employee Cooperation and Awareness

In-place preservation heavily relies on the cooperation and awareness of employees who are custodians of potentially relevant data. Organizations should ensure that employees understand their obligations, provide clear instructions, and have mechanisms to address non-compliance or lack of awareness.

Legal and Regulatory Compliance

In-place preservation should align with applicable legal and regulatory requirements specific to the jurisdiction and industry. Organizations should consult legal professionals to ensure compliance with relevant laws and regulations while implementing legal holds.

Monitoring and Legal Holds Management

In the current age of litigation, custodians’ data is not limited to email or file-sharing, or collaboration platforms; it could be cell phone data, social media data, or IM (Instant Messaging) applications like WhatsApp, Snapchat, etc.

These applications and devices may not support In-place preservation at the moment. In this scenario, the legal team has to rely on conventional legal holds technique for preserving data from all the sources that don’t support in-place preservation. This could lead to a sailing on two boats situation for legal teams.

Wrapping up

It is important for organizations to carefully evaluate these caveats and develop comprehensive strategies to address them when implementing in-place legal holds. Seeking guidance from legal and technology experts can help ensure the effectiveness and defensibility of the preservation process.


AI in eDiscovery: A Beginner’s Guide

Electronic discovery/disclosure is an integral part of the dispute resolution process…

eDiscovery pertains to identifying, collecting, reviewing, and producing electronically stored information (ESI) in response to a lawsuit or investigation for example. For the most part lawyers using eDiscovery are already under time pressure and just want to get on with reviewing the document batches set up in the system. But the high-tech facets are becoming more and more essential as the volume of the data grows, so it is only practical to pay attention to AI from now on. AI features such as Technology Assisted Review (TAR) bring efficiency and order to evidence gathering, reducing the time required to sift through physical documents.

Compared to the pre-digital age manual methods, eDiscovery is a vastly improved way to gather and maintain evidence, but it is now facing challenges with among other things, the sheer amount of ESI sourced from emails, text messages, social media, cloud-stored files, etc. The volumes can be overwhelming especially when we consider that legal experts must take time to identify, collect, and review this information, comply with legal and regulatory regulations, and ensure that data relevant to the case is preserved and provided with strong security measures that protect privacy.

For all its advantages, today’s eDiscovery still takes tremendous effort and time. According to a 2022 survey, the greatest efforts for core eDiscovery tasks (i.e., processing, collection, and review) occur during the review phase, with nearly 66% of expenditures used by this one task alone. So, any drive at efficiency should really focus on the review stage. ( As the types and volumes of ESI skyrocket, the conventional approaches only get more expensive and time-consuming. We are now at the point where legal teams must consider faster, more effective ways sifting through the data.

Artificial Intelligence (AI)-powered eDiscovery is one possible solution…

As the name implies, AI-powered eDiscovery software leverages cutting-edge artificial intelligence technologies such as Machine Learning (ML) and Natural Language Processing (NLP) to accelerate and enhance the identification, processing, review, relevancy, and even redaction of documents produced in court and shared with opposing counsel.

AI promises several benefits over the current eDiscovery methods that result not only in faster analysis among vast amounts of data but also improved accuracy and review consistency, enhanced categorization capabilities, prioritization through predictive coding (like TAR where a computer uses AI to identify and tag potentially relevant documents), pattern and trend identification often missed by the human eye, and, of course, the bottom-line savings in cost and resources. Please see

AI certainly holds a lot of promise, but it’s essential to approach it with a balanced view. Data experts caution that we should separate the hype from reality, keeping the current limitations of AI in mind. While AI excels in generating creative ideas and possibilities, it struggles with tasks involving real data and concrete observations. AI can be brilliant at creating images or writing poetry, but its interpretation of factual learned information from its training database can sometimes be less impressive.

This becomes particularly problematic in specificity areas like legal interpretations. AI’s shortcomings become evident when precision and factual accuracy are paramount. AI tools like Chat-GPT for example can be prone to conveying misinformation with an unwarranted degree of certainty and authority. Therefore, it’s important to remember that despite its potential, AI is still a developing tool. Its outputs should be carefully evaluated and cross-verified, especially in a sensitive domain like the law.

Current challenges we are seeing in the application of AI in a legal context…

A famous example of AI’s occasional lapses in logical reasoning can be seen when testing OpenAI’s Chat-GPT 3.5:

  1. When posed with the question, “If a woman can make one baby in 9 months; how many babies can 9 women make in one month”? the AI delivered a mathematical answer saying nine women can collectively make one baby in one month, without accounting for the biological reality.OpenAI's Chat-GPT 3.5 - "baby 9 month" Prompt
  2. The newest iteration, Chat-GPT4, provides a more accurate response essentially saying that, ‘the process of human gestation does not increase in speed with the number of women involved’. This output shows that successive generations of AI transformers are improving and grasping reality better.

But the above example serves as a reminder of how even simple logic can be distorted when AI algorithms misinterpret information. It again raises the question of whether we can trust AI enough to be a reliable tool in eDiscovery. Does the hype live up to the reality? The answer, for now, is, “It depends.” Utilizing AI tools for legal advice without human oversight carries significant risks. AI can be helpful for straightforward tasks such as drafting non-disclosure agreements or composing traffic violation protest letters, which are easy to template and cross-check.

However, more intricate legal tasks should involve human legal expertise and plenty of overview. This does not mean that the promises of AI and eDiscovery benefits do not have future potential, on the contrary. With time, patience, and further technological advancements, we may ultimately reach a point where AI and legal expertise start to intersect. Already there are techie discussions about how ChatGPT and similar tools can participate in eDiscovery with relevant synonym searches on online queries, text generation of response letters, legal chatbots for simple FAQs, and creating blogs and social media posts.

Yet, we must remember that this technology is still in its developmental stages and has limitations. Moreover, the utilization of AI in eDiscovery differs significantly from other applications. Lawyers engaged in eDiscovery are not merely seeking answers but are searching for evidence to support their case. The law et alia, will provide a great “track and field” to test the boundaries of AI.

Legal analysis requires a certain finesse and skill that only human expertise can provide to the final degree. Even if we employ AI tools to present two sides of an argument, it’s the role of legal professionals to interpret and apply this information. This scenario further underscores the fundamental role of human involvement in law, highlighting that we cannot (nor should we try to!) replace human expertise with AI. These days tech companies often include a quality control clause in their employee handbook which essentially says that workers may use AI and third-party model builders if they wish, but ultimately the responsibility for the work product remains with the person, so human verification is essential.

In conclusion, AI is potent, but it requires careful handling, scepticism, and a thorough understanding of its capabilities and limitations. AI-powered eDiscovery is a reality, already in terms of its current applications and proven benefits. As we continuously refine this technology, integrating AI and legal expertise will progressively pave the way for the future of automated eDiscovery. Watch this space!