Lead story
Deleted AI prompt records lead to federal discovery sanctions
A Northern District of California court found spoliation after defendants deleted prompt-engineering files and other ESI in a trade-secret case.
- What happened
- A Northern District of California court found spoliation after defendants deleted prompt-engineering files and other ESI in a trade-secret case.
- Why it matters
- AI evidence may reside outside the tool that created it, including cloud drives, development systems, and local devices. Litigation holds, custodian interviews, and source maps should cover prompts, outputs, logs, copied exports, and the repositories to which users move them.
- Watch
- The underlying spoliation order was issued September 17, 2026; the current source is a November 18 legal analysis of that order. The parties later settled, and the court did not find the intent required for Rule 37(e)(2) sanctions.
Sourceshaystackid.com
Briefing item
Personal-device discovery turns on employer control
A New York State Bar Association analysis explains the competing tests courts use to decide whether employers control work-related ESI on employees' personal devices.
- What happened
- A New York State Bar Association analysis explains the competing tests courts use to decide whether employers control work-related ESI on employees' personal devices.
- Why it matters
- Whether a company must preserve and produce personal-device data can depend on policy language, actual business practices, and the jurisdiction's control standard. Organizations should align device policies, employee consent, and hold procedures before a dispute makes those facts outcome-determinative.
- Watch
- The article synthesizes decisions from several federal districts; it does not describe a uniform national test. The cited cases turn on distinct policies, usage patterns, and evidence of access.
Sourcesnysba.org
Briefing item
Courts reject a categorical AI privilege-waiver rule
A Bloomberg Law analysis finds that recent courts are treating AI tools as instruments of legal work rather than automatic third-party waivers of work-product protection.
- What happened
- A Bloomberg Law analysis finds that recent courts are treating AI tools as instruments of legal work rather than automatic third-party waivers of work-product protection.
- Why it matters
- Privilege review for AI material should separate the protected purpose and content from the platform's handling of data. Counsel should document tool tier, retention settings, contractual confidentiality, training restrictions, inputs, and reviewer direction instead of assuming that all AI use either preserves or destroys protection.
- Watch
- This is practitioner analysis of developing, fact-specific authority, not a categorical holding applicable in every jurisdiction. The article was corrected to refine its descriptions of several decisions.
Sourcesnews.bloomberglaw.com
Briefing item
Rajasthan court orders preservation of vulnerable call records
The Rajasthan High Court held that a fair trial includes a meaningful opportunity to preserve potentially relevant electronic evidence before routine deletion makes it unavailable.
- What happened
- The Rajasthan High Court held that a fair trial includes a meaningful opportunity to preserve potentially relevant electronic evidence before routine deletion makes it unavailable.
- Why it matters
- Time-sensitive telecom and platform records may disappear before the ordinary evidence phase. Parties seeking such material should identify retention limits early and frame preservation as maintaining availability, separately from later disputes over admissibility and weight.
- Watch
- The ruling concerns Indian criminal procedure and an accused person's fair-trial rights. LiveLaw's report was reviewed, but the underlying order was not independently accessible from the source page.
Sourceslivelaw.in
Briefing item
China's AI-disputes guidance sets proof and recordkeeping expectations
New analysis of China's first comprehensive judicial policy for AI disputes highlights evidentiary burdens for developers, providers, users, and rights holders.
- What happened
- New analysis of China's first comprehensive judicial policy for AI disputes highlights evidentiary burdens for developers, providers, users, and rights holders.
- Why it matters
- AI disputes increasingly depend on technical records that may be distributed across model development, deployment, testing, and notice-response systems. Teams operating in China should map those records to likely proof burdens and maintain context sufficient to reproduce challenged outputs.
- Watch
- The Supreme People's Court issued the Opinion on September 7, 2026; the current item is a November 17 analysis, not a new issuance. The linked bilingual text is an unofficial AI-generated and human-edited translation, so the Chinese original controls.
SourcesJD Supraloeb.com
Educational only. Not legal advice. Review sources and matter-specific obligations before relying on any workflow decision.