USA - Nationwide: An E-Discovery & Information Governance Overview

Courts and companies are grappling with the accelerating complexity of electronically stored information (ESI) and the disruptive effects of generative artificial intelligence (AI) on eDiscovery and information governance. Developments reflected in recent judicial decisions reinforce long-standing core issues – preservation, proportionality, privilege and protecting confidentiality – while also clarifying how those principles apply in a technology environment that looks very different from that of even a few years ago.

Information Governance and eDiscovery Processes: Risk Multipliers

Deficiencies in information governance and eDiscovery processes are increasingly not just background compliance issues but also a direct driver of discovery failures and sanctions exposure. Judicial opinions increasingly trace discovery breakdowns to systemic gaps: failure to align retention practices with legal obligations; failure to operationalize legal holds to deal with new sources of relevant information resulting from new tools and technologies; and over-delegation of relevance and privilege judgments to unsupervised custodians.

A recurring theme is that failures are rarely isolated mistakes. Courts are willing to characterize patterns of late production, over-designation of information as privileged, and failures to preserve relevant information as persistent and systemic, and to attribute responsibility not only to organizations but also to counsel who fail to exercise meaningful oversight. Monetary sanctions – sometimes substantial and punitive – are being used not merely to compensate the opposing party but also to enforce discipline where lesser remedies are viewed as insufficient.

Preservation Duties Continue to Expand Well Beyond Email

Courts remain focused on preservation failures involving non-email data sources. Messaging platforms, collaboration tools, mobile devices, ephemeral communications, structured data and hard-copy materials all feature prominently in sanctions decisions. Issuing a legal hold is no longer treated as a safe harbor if the organization lacks the technical or operational capacity to actually preserve the data subject to the hold.

Importantly, the risks of both under- and over-preservation continue to increase. While failures to preserve relevant data often result in sanctions, indiscriminate “save everything” approaches are particularly problematic due to vastly expanding data volumes. This approach magnifies IT burdens and costs, privilege risk, privacy exposure, cybersecurity vulnerability, and review volumes and complexity. Proportional and defensible preservation – grounded in an understanding of systems, data flows and business practice – is becoming a necessity rather than a “nice to have” best practice.

Discovery Failures: Greater Consequences for Litigants and Counsel

Another strong trend is judicial willingness to impose severe remedies and even punitive sanctions when discovery obligations are treated casually. Courts are increasingly explicit that compelled “disclosure after exposure” of deficiencies does not cure prejudice, and that late or corrective productions do not erase the underlying misconduct or its impact on the integrity of the process.

Equally notable is the scrutiny applied to counsel’s role. Courts are emphasizing that delegation of critical discovery decisions – such as relevance, collection scope and privilege determinations – to non-lawyers without meaningful oversight is incompatible with professional responsibility. Credibility erosion following over-withholding or inaccurate privilege assertions is frequently an additional adverse consequence.

Discovery Materials’ Confidentiality Gains Renewed Focus

Courts are confirming that discovery materials remain non-public, even without a formal confidential designation. In particular, courts are pushing back against arguments that discovery materials should be deemed effectively public – and therefore subject to public disclosure and sharing – simply because they do not meet formal protective order confidentiality designation requirements.

Relatedly, courts are beginning to address whether produced materials can be input into AI models, particularly public generative AI platforms. Inputting discovery materials into public or open AI platforms – particularly those operating under terms that permit reuse, training or disclosure of the data – is increasingly seen as inappropriate. By contrast, closed or enterprise AI tools with contractual confidentiality safeguards, defined data boundaries, auditing and supervision are viewed as presenting materially different risk profiles.

Judicial concern has centered on the irreversibility of inputting material into public AI platforms, where the information may become publicly accessible. Once uploaded into an open or public generative AI platform, information may be retained, incorporated into data that trains the model, and disclosed directly or indirectly at scale in ways that cannot be clawed back or meaningfully remediated. Courts have treated this risk itself as sufficient good cause to bar submission of discovery materials into public AI platforms.

Heightened Scrutiny of AI-Generated Court Submissions

Courts are signaling zero tolerance for misuse of generative AI in filings. Early judicial encounters focused on fabricated case citations. More recent decisions also address fabricated or distorted facts presented under oath, often in declarations or affidavits filed at dispositive stages.

Sanctions imposed in this context reflect two linked expectations: that attorneys must independently verify all factual and legal assertions regardless of the tools used, and that clients and counsel share responsibility for sworn submissions. Courts have shown particular frustration where parties attempt to deflect fault to technology rather than acknowledging and correcting errors promptly.

The consistent message is that AI does not dilute counsel’s duties of candor, diligence or supervision. Its use heightens – not lessens – the expectation of robust quality control and validation processes.

Privilege Under Pressure: Mixed-Purpose Communications and Heightened Scrutiny

Privilege review is becoming more challenging as communications with mixed legal and business purposes become the norm rather than the exception. At the same time, parties are increasingly aggressive in challenging opponents’ privilege claims, and courts are more frequently rejecting assertions of privilege where the basis for protection is not clear and evident. Courts are also showing growing intolerance for over-designation and delayed productions due to privilege re-reviews and downgrades, emphasizing the need to get privilege review right the first time around.

Emerging Tensions Regarding AI Use and Privilege

With respect to whether AI inputs and outputs may be a privileged or protected work product, an emerging framework treats AI as a drafting or research tool akin to other technologies, rejecting the notion that its use inherently constitutes disclosure to a third party. Another framework focuses on the confidentiality expectations created by platform terms and holds that use of certain tools – ie, open or public AI tools – may undermine claims of protection. Outcomes may turn on whether materials were created because of anticipated litigation, whether counsel was meaningfully involved, and whether reasonable steps were taken to preserve confidentiality.

Courts Expect Early, Informed Engagement Regarding Technology and Process

Across these developments, courts consistently emphasize the value of early engagement. Judges expect parties to identify relevant data sources promptly and thoroughly, to negotiate ESI protocols thoughtfully, and to address AI use explicitly rather than implicitly. Courts are increasingly willing to characterize late-breaking disputes as avoidable – and therefore potentially sanctionable – when they stem from a failure to engage early with known technological or governance issues.

Conclusion

These trends reflect an environment that is becoming increasingly demanding regarding eDiscovery and information governance. Courts are insisting that innovation be matched by governance, transparency, competence and accountability. Organizations that invest in mature information governance, clear AI policies, disciplined eDiscovery processes, and early, technically informed legal strategies are best positioned to navigate this environment.