AI intellectual property attorney time tracking: USPTO AI examination guidance, copyright authorship of AI-generated works, and trade secret AI training data advisory calls

USPTO AI examination guidance: calls on the Patent Office's examination schedule

The USPTO's February 2024 "Inventorship Guidance for AI-Assisted Inventions" (89 Fed. Reg. 10043) establishes that AI systems cannot be named inventors — only natural persons who make a significant contribution to the conception of the claimed invention under the Pannu factors satisfy the inventorship standard. The July 2024 updated "2024 USPTO Guidance on Patent Subject Matter Eligibility" addresses AI/software patent claims under Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014), and the revised Step 2A Prong Two framework. When a USPTO examiner issues an Office Action rejecting an AI-assisted invention claim under these guidelines — finding that the human contribution is insufficient to constitute inventorship, or that the claimed AI-implemented method fails Step 2A Prong Two — the AI patent prosecution attorney initiates advisory calls with the client's engineering team, the USPTO examiner, and the client's in-house IP committee. These calls arrive on external schedules that are entirely independent of the patent prosecution attorney's billing calendar.

Five USPTO AI examination call types: (1) client's AI engineering team call to document human inventive contribution for each claim element (30–45 min) — arrives on the engineering team's sprint calendar when the team has bandwidth to review the Office Action's claim-by-claim rejection and identify the specific human contributions to each element; (2) USPTO Applicant-Initiated Interview (AII) preparation call with co-inventor to confirm human contribution statement accuracy (20–30 min) — arrives when the AII is scheduled on the USPTO examiner's availability calendar; (3) § 101 Alice Step 2A claim amendment strategy call with in-house patent counsel (25–40 min) — arrives when the claim amendment strategy requires internal IP committee approval on the client's governance calendar; (4) continuation claim scoping call with the AI product team (20–30 min) — arrives on the product roadmap schedule as the AI model releases new capabilities that may support broader continuation claims; (5) patent portfolio audit call to identify AI-assisted inventions requiring inventorship correction under 37 C.F.R. § 1.48 (25–35 min) — arrives when the client's IP audit is scheduled on the client's fiscal year calendar. At 55% untracked: 10 AI patent prosecution matters × 5 calls × 35 min × 55% = 16.04 hrs ≈ 16.0 hrs = $7,200–$12,000/year at $450–$750/hr.

The inventorship correction and portfolio audit calls are the most frequently untracked because they arrive when the client's IP audit schedule reaches the AI-assisted inventions portfolio — a date entirely within the client's fiscal year control and disconnected from any USPTO filing deadline or examination window. AI patent prosecution attorneys managing active prosecution files are also handling portfolio-wide correction obligations that arrive on a client-driven schedule, compounding the total billing gap beyond what prosecution-only matter calendars capture.

Copyright authorship of AI-generated works: calls on the Copyright Office's processing schedule

The Copyright Office's February 2023 Zarya of the Dawn decision (partial cancellation of AI-generated image elements), August 2023 "Copyright and Artificial Intelligence" Part 1 notice of inquiry, and March 2024 Copyright Office Policy Statement establishing that "works produced by human beings" must contain sufficient human authorship for copyright registration eligibility have created an active advisory practice for AI intellectual property attorneys. When the Copyright Office issues a correspondence letter questioning the human authorship of a registered work — triggering a registration cancellation proceeding — or rejects a new AI-assisted works registration application, the AI intellectual property attorney advises the client on registration strategy and human authorship documentation. These advisory calls arrive on the Copyright Office's examination processing queue — currently 3–9 months for most registrations — and on the client's creative team's project schedule, neither of which is visible on the attorney's matter management calendar.

Four copyright authorship advisory call types: (1) Copyright Office correspondence response advisory call with the client's creative team (25–40 min) — arrives when the Copyright Office's Registration Examiner issues a correspondence letter questioning human authorship on the Copyright Office's processing schedule, triggering an immediate call to assess the registration strategy; (2) human authorship documentation strategy call reviewing creative briefs, prompt drafting logs, and selection and arrangement records (20–35 min) — arrives when the client's creative director is available on the creative team's project schedule to gather the documentation needed to respond to the Copyright Office's correspondence; (3) copyright portfolio audit call identifying AI-assisted works requiring authorship re-analysis under the Office's 2024 policy (25–35 min) — arrives when the client's IP audit schedule reaches the copyright portfolio on the client's fiscal year calendar; (4) licensing advisory call for AI-generated works where copyright subsistence is uncertain (20–30 min) — arrives on the client's licensing deal schedule when a prospective licensee raises copyright subsistence concerns before executing the license. At 55% untracked: 8 matters × 4 calls × 30 min × 55% = 8.8 hrs = $3,960–$6,600/year at $450–$750/hr.

The licensing advisory calls are structurally distinct from the registration advisory calls: they arrive not on the Copyright Office's processing schedule but on the licensee's deal schedule — when a prospective licensee's due diligence review surfaces questions about AI-generated content's copyright status and the licensing deal timeline requires immediate legal advice. Standard matter-based billing platforms that track copyright prosecution timelines cannot surface calls that arrive when a third-party licensee's counsel raises a copyright subsistence question on the licensee's deal closing schedule.

Trade secret AI training data: advisory calls during litigation discovery

When a competitor allegedly uses proprietary AI training data — annotated datasets, fine-tuning corpora, proprietary model weights, customer datasets assembled at significant cost — to develop a competing AI model, the trade secret plaintiff's attorney advises on whether the training data constitutes a protectable trade secret under the Defend Trade Secrets Act (DTSA), 18 U.S.C. § 1836, and applicable state law including the California Uniform Trade Secrets Act, Cal. Civ. Code § 3426 et seq. The plaintiff must establish that the training data derives independent economic value from not being generally known, and that the owner took reasonable measures to maintain its secrecy. Advisory calls during the litigation discovery period arrive on five different external schedules, each driven by a different party's operational calendar.

Five DTSA AI training data advisory call types: (1) client's AI engineering team model comparison briefing call (30–45 min) — arrives on the engineering team's sprint cycle when the team completes its analysis of the competing model's outputs, model card specifications, and benchmark performance metrics to evaluate whether the competing model may have been trained on the plaintiff's proprietary data; (2) technical expert engagement and scope call (25–35 min) — arrives when the technical expert has availability on the expert's consulting schedule to discuss the engagement scope and analysis methodology; (3) DTSA inevitable disclosure doctrine advisory call for departing AI engineer departure risk (20–30 min) — arrives when the HR team identifies a departure risk on the HR team's personnel management calendar; (4) Computer Fraud and Abuse Act (CFAA) and hiQ Labs v. LinkedIn Corp., 31 F.4th 1180 (9th Cir. 2022), data scraping advisory call (25–35 min) — arrives when the litigation discovery process surfaces evidence of scraping activity on the opposing party's document production timeline; (5) stipulated protective order model weight production negotiation call with opposing counsel (20–30 min) — arrives on opposing counsel's briefing timeline when the scope of model weight production under the protective order requires negotiation. At 55% untracked: 6 matters × 5 calls × 28 min × 55% = 7.7 hrs = $3,465–$5,775/year at $450–$750/hr.

The inevitable disclosure advisory calls for departing AI engineers arrive with particular unpredictability: the HR team identifies the departure risk on the HR team's internal personnel management calendar, triggering an immediate call to the AI intellectual property attorney for advice on whether to seek a temporary restraining order or pursue a covenant not to compete — both of which are time-sensitive enough that standard billing timers are rarely running when the call begins.

How ClaimHour fits AI intellectual property practice

If you prosecute AI-assisted inventions under the USPTO's 2024 inventorship and subject matter eligibility guidance, advise on copyright authorship of AI-generated works under the Copyright Office's 2024 Policy Statement, and litigate trade secret misappropriation of proprietary AI training data under the DTSA — and your invoices consistently understate the engineering team sprint-calendar calls arriving outside USPTO examination windows, the creative team project-schedule calls arriving outside the Copyright Office's registration processing queue, and the HR-triggered departing-engineer calls arriving with no advance notice — ClaimHour was built for that gap. The passive capture logs every client call (iOS call metadata: duration, timestamp, direction — not content), every email advisory session, and every document review session. A 2-minute evening digest surfaces each unmatched call for matter attribution. No audio. No call contents. No email bodies. Privilege is preserved under ABA Formal Opinion 512. Join the waitlist and we'll email when early access opens.

Get early access

Related questions

Further reading

• Intellectual property attorney time tracking — traditional patent prosecution, licensing, and IP litigation billing gaps; the companion page covering the broader IP practice billing gap structure that underlies the AI-specific USPTO examination and copyright authorship framework covered here
• Patent prosecution attorney time tracking — USPTO examination coordination billing gaps for traditional patent prosecution; structurally parallel to the AI patent examination call gap but without the engineering-team sprint calendar dimension that characterizes AI-assisted invention prosecution
• Trademark attorney time tracking — USPTO trademark examination calls and brand protection advisory for AI-generated branding and AI-assisted logo design; the trademark companion covering the copyright and trademark overlap in AI-generated brand assets
• Cybersecurity attorney time tracking — AI model security incident response advisory calls and NIST AI Risk Management Framework (AI RMF) compliance advisory; the cybersecurity companion covering the security and privacy dimension of AI model deployment that generates parallel advisory calls on the incident response team's schedule
• Data privacy attorney time tracking — AI training data subject rights under CCPA/CPRA § 1798.121 and EU AI Act compliance advisory calls; the data privacy companion covering the consumer rights dimension of proprietary AI training data that generates parallel advisory calls on the privacy compliance team's schedule
• Startup counsel attorney time tracking — early-stage AI startups facing IP ownership disputes with founding engineers and AI IP assignment clause advisory; the startup companion covering the employment and equity context in which AI training data and AI-assisted inventions are most frequently contested at formation and Series A