Who Owns the Home Office? Copyright and Contractual Control in Remote Work

Introduction : Picture this. A marketing staff member writes a presentation for a new product launch that takes him 3 weeks to create. She uses a personal laptop, uploads drafts to a personal Google Drive folder, uses an AI writing assistant to review drafts and sends versions to colleagues for their feedback via WhatsApp. Six months later, she starts working for a competitor. Now, her previous employer claims the ownership of the presentation. But, she argues that it was done by her on her own time, on her own device, and using her own tools.

What is the ownership of it? And, above all - who can prove it?

The real question is, not ownership, but control. The copyright dispute has become from who owns a work to who can identify, preserve, prove, and enforce ownership as content flows easily between personal devices, cloud platforms, messaging apps, and AI tools. This is a quite different problem and requires a quite different solution.

Relevant Legal Provisions

The following statutes and principles are relevant to this discussion:

Copyright Act, 1957 (India)

Section 2(d): Restricts authorship to natural persons, and does not explicitly recognise AI systems as authors of generated works.

Section 13: Copyright applies to original literary and artistic works, including the content, presentations, reports, templates, and written works created during remote work.

Section 17: The main ownership provision. Works produced by an employee in the course of his/her employment under a contract of service are presumed to be vested in the employer, unless otherwise agreed.

Section 18: Defines copyright assignment - the basis on which IP assignment clauses in employment contracts operate.

The exceptions to infringement under Section 52: Lists include exceptions for research and private study. Relevant to the permissible use of AI-generated content.

The Copyright (Amendment) Act, 2012 - included provisions on digital content and internet use.

Bharatiya Sakshya Adhiniyam, 2023 - In place of the old Section 65B of IEA, electronic records must now have a certificate of authenticity that will be admissible in court. It is essential in any metadata, version history, or digital audit trail dispute.

The Home Office as a New IP Battlefield

From an IP point of view, the traditional work environment was more of a controlled environment. They used company hardware, saved to the company servers, and communicated via company monitored systems. The physical and digital perimeters were more or less the same. In the event of an argument over the origin of a manuscript, the answer was typically known.

This structure has been broken down by remote working. Workplace is no longer a place. It’s a network, a personal network, and an invisible network to the employer.

An employee can write a report on their MacBook, edit it on their phone on the way to work, use a free version of an AI assistant to fine-tune the report, save an early version to iCloud, and a final version to a company SharePoint, and then discuss changes on Signal. All of these activities create a digital trail in some way. Very few of them leave a footprint in a place that the employer controls.

There are three consequences. Firstly, there is no clear distinction between personal and work time, and personal devices become work devices. Second, evidence is spread out in platforms, jurisdictions, and accounts that are inaccessible to employers. Third, even if legal rights to it are secured, it will be hard to enforce as the practical means of obtaining it, keeping track of it, and limiting access to it are spread across systems that are not under company control.

This is the battlefield. But the question remains, has most employment contracts - and most HR policies - caught up with it?

Ownership Is Usually Clearer Than We Think

This is where the majority of the commentary about this subject goes wrong. In most cases, ownership issues are not the difficult issue.

The general principle of the Copyright Act, 1957 is that the first owner of a work is the author of the work under Section 17 of the Copyright Act, 1957. Where, however, a work is created by an employee in the course of the employment (under a contract of service), the employer is - in the absence of a contract to the contrary - regarded as the first owner of the copyright. It is a settled fact in Indian law which has been reiterated in various judicial situations. The principle is simple: When you’re a content creator, presenter, report writer, or creative asset generator, you’re typically working for your employer, so the resulting work is theirs.

The position is different for independent contractors. The copyright of a freelancer who is employed under a contract for services belongs to him unless the contract contains a specific provision to the contrary. Many conflicts have arisen over this distinction, especially in the gig economy world, and it is important to have a clear definition of the contractual language.

When organisations have well drafted clauses regarding the assignment of work product in their employment contracts, including clauses which clearly define what constitutes work product, assign ownership of all work product to the employer, and address future works, the issue of ownership is generally answerable with reasonable confidence.

In Eastern Book Company v. D.B. Modak, the Supreme Court of India had determined what level of originality is necessary for copyright protection. The Court explained that a work must show the skill, judgment, and intellectual contribution of the author to be protectable, and that something which is merely mechanically reproduced is not protectable. In the world of telework, the principle translates into a practical reality: Routine or formulaic output (like a report generated by computer) may not be protected by copyright, while documents that reflect an employee’s original intellectual contribution (like a draft, analytical framework, strategic presentation) will be. This is important if an employer wants to claim ownership of a broad range of employee-created content.

The legal landscape, therefore, is somewhat stable. But the case is not same in practice.

The Real Problem: Control Without Possession

In the traditional copyright ownership, there is an element of possession. You are the owner of the asset and can collect it, block it, and assert your rights to it. The connection between remote work and the job has been broken.

For the employer who owns the copyright to the employee’s training manuals, pitch decks, brand guidelines, and internal reports, but doesn’t know where they are, how many copies exist, who has access to them, or if they have been shared with an external party. Legal ownership persists. There is no practical control.

This has been exacerbated by the Bring Your Own Device (BYOD) culture that has increased exponentially during the pandemic. Where employees use their own laptops and phones for work, the employer’s access, monitoring and audit rights are legally limited - not always by the employment policies and data protection policies that were intended to safeguard the privacy of the employee. The employer owns the copyright to the work, but may not have access to the device that the work was created on.

The problem is exacerbated by cloud storage. The files may be saved in a Google Drive or Dropbox account that is not the same as the one the employee is using for work, and may have been created before or in parallel with the employee’s work account, making it very hard to determine which versions are for which purpose. Personal and professional folders are a click apart.

Messaging adds another layer of complexity. Discussions, drafting or revision of work on WhatsApp, Telegram or Signal is on the personal device and in accounts that are not under the employer’s control. Remote work has created a world where copyright ownership can exist even without practical control.

The Forgotten Asset: Drafts, Templates, and Work-in-Progress

The discussion of intellectual property rights in employment relationships tends to be about the end product - the presentation, the article, the product. But there is little focus on what may well be the more commercially relevant group: work-in-progress.

Drafts and templates can be much more useful in knowledge-intensive industries than the final product. The template reflects the organisation’s way of thinking, structure and institutional jargon. The draft captures the creative process - how a team frames a problem, how they talk about the products, how they analyze a problem. An employee who leaves with these assets isn’t just taking the information they contain. They take the methodology.

The Copyright Act, 1957, Section 17 does not differentiate between complete and incomplete works. The same rules for ownership apply. However, the reality for drafts is even more difficult: drafts are more likely to be in personal folders, more likely to have been edited across devices and less likely to be subject to specific contractual provisions that were drafted with final deliverables in mind.

Organisations need to ask themselves a tougher question than, “what do our employees produce?” They should ask: “What thinking, structure and institutional knowledge do our employees have - and how has it been identified and preserved?”

Copyright Leakage vs. Information Leakage

While there is a long history of separation between confidentiality obligations and IP ownership in employment law practice, the two are often confused in risk discussions. They should not be.

Imagine an employee who copies a presentation and a few internal templates to his personal drive when he leaves the company. For purposes of this discussion, let’s say there is no confidential information, no trade secrets, no customer information, no commercially sensitive financial information. There confidentiality concern is limited but the copyright concern is not.

The employer may still be the owner of the copyright in those works. Unauthorised copying of any material, whether or not it is confidential, may be a violation. More significantly, the reuse or adaptation of those works, by the employee in a new role, as a freelancer or as a new enterprise, may continue to infringe even after the confidential information it once contained has been made public or become commercially irrelevant.

Confidentiality training is only a significant part of the risk if it’s the only training that’s done. There can be copyright leakage without the knowledge of the copyright owner, without the appearance of any law, and without any violation of any non-disclosure agreement - and it can still be a real commercial damage.

The Emerging Evidentiary Crisis

How will it be determined who owns it if it is contested?

In many respects this is the most important question in employment IP law today and in practice it is being answered by digital forensics, not legal doctrine.

The first line of evidence is metadata. Digital files include information that is embedded in the files, such as creation date, modification history, author names, and editing time. According to Indian law, metadata attached to digital files can be used as evidence in court proceedings, subject to the requirements of the Bharatiya Sakshya Adhiniyam, 2023 (the Indian Evidence Act, 1872 has been replaced). Most importantly, electronic records must be accompanied by a Section 65B equivalent certificate of authenticity to be admissible as evidence in court cases. Even probative digital evidence can be suppressed without such certification.

Version history is a crucial factor in collaborative tools such as Google Docs, Microsoft 365, or Notion. It audit logs in enterprise systems like SharePoint or Confluence, provides time stamped trails of every edit, and is most useful when preserved proactively, not retrieved after a dispute has solidified.

AI content could very well be the biggest evidentiary hurdle of the next decade. If a human worker is using AI to draft a document, the evidence of the AI’s work is scattered: The AI model has created parts of the document, the human worker has prompted and directed the AI, and the document may contain no evidence of where it came from. As per the existing law, the Copyright Act, 1957, non-human authorship is not recognised in India as authorship is defined in Section 2(d) of the Copyright Act, 1957 to be limited to natural persons. Works assisted by AI with significant human creative input might be eligible for protection, but the challenge of proving that the human contribution is increasingly difficult to distinguish.

The party who has the stronger legal argument may not win the copyright battle. It could be won by the party that has the more superior metadata.

AI Has Made the Home Office Even Harder to Define

The AI layer is a unique one in that it introduces risks on several fronts, though it doesn’t alter any of the core tenets of ownership.

There are two things that occur when an employee uploads company content to an AI platform and it generates a report. For one, the information is likely to be transferred from the employer’s controlled environment into the training or processing system of a third party service. Second, the output is co-authored in a manner that is unlikely to have been foreseen by the pre-existing contracts.

Who is the owner of the AI-generated draft? Where the employee was encouraged to use company strategy documents, competitive intelligence or internal frameworks, and the output from the session is a result of that institutional knowledge, it would seem that the employer has a strong claim, even if the employee did it on his own time.

In ANI Media Pvt Ltd v. Open AI Inc & Anr, ANI, a news agency, had sued OpenAI for allegedly scraping and using its copyrighted news articles without permission or payment. The case asks two direct questions regarding the context of remote work: whether the use of third party copyrighted material for the purposes of training or processing an AI system will be considered infringement under the Copyright Act, 1957; and whether the exceptions in this Act under Section 52 would apply to the training of an AI system. Interim rulings have recognized the complexity of AI-generated content and the lack of a clear legislative framework, although the Court has yet to render a final decision. The case is a warning for employers that employees who upload internal documents for the company that contain proprietary information to AI platforms could be exposing themselves to liability in more ways than just data leakage - potentially for committing to acts that courts could later find infringing.

Meanwhile, employees are employing AI tools that their employment contracts do not cover, on devices that their employment contracts do not cover, on platforms that their employment contracts do not cover. It’s a significant governance deficit.

From Ownership to Governance: What Employers Should Actually Do

The above problem can’t be solved using generic recommendations like “update your contracts” or “train your employees”. A governance framework that spans four layers is needed.

Layer 1 - Precise Copyright Clauses

Employment contracts should go beyond the standard ‘work product is the property of the employer’ clause. Work product should be defined as drafts, templates, derivative works, works created with the use of AI, works created with personal devices or personal accounts in connection to employment duties. The meaning of the “course of employment” language found in Section 17 of the Copyright Act, 1957, is not always as straightforward as it might be, and a clear contractual definition would help avoid ambiguity in the event of a dispute.

Layer 2 - Device and Platform Governance

BYOD policies should not simply recognise the use of personal devices. They should define where employees are allowed to post work-related information, if employee-owned cloud accounts can be used to store employer IP, and what happens to that information upon termination.

Layer 3 - Evidence Preservation by Design

This is probably the least developed aspect. Organisations should set up their systems to create and maintain audit trails in collaborative tools, access logs in document management systems, and metadata retention policies.

Layer 4 - AI Usage Policies

With the rate of adoption, the use of AI is no longer a choice, but a requirement. These should include details on the AI tools that are allowed to be used for work-related activities, the prohibition on uploading confidential or proprietary data to non-approved AI platforms, the ownership of the AI-generated output, and the requirement for disclosure when AI is used to create content that will be presented as the employee’s work product.

Conclusion

The issue of ownership of content generated in remote work setting is no longer just a matter of whether employers own the content. In many situations, this is answered by the law already - Section 17 of the Copyright Act, 1957 establishes the principles of ownership for employed workers, and in most other situations the ambiguities can be resolved through well drafted assignment clauses.

But the more challenging question is whether employers can detect it, save the evidence of its production, and enforce rights over it when it has spread across personal devices, personal accounts, personal messaging apps and AI platforms that are outside of the scope of any corporate IT policy.

The law of copyright hasn’t changed, regardless of remote work. It has transformed the context of copyright functioning - and now practical control of intellectual assets is just as significant as legal ownership of them. The home office is likely to hold more unprotected IP than organisations realise, especially if they haven’t yet considered copyright governance as an operational function.

Author: Jyotsana Singh, in case of any queries please contact/write back to us via email to chhavi@khuranaandkhurana.com or at  Khurana & Khurana, Advocates and IP Attorney.

References and Sources

1. Copyright Act 1957 (India) ss 2(d), 13, 17, 18, 52.

2. Copyright (Amendment) Act 2012 (India).

3. Bharatiya Sakshya Adhiniyam 2023 (India) s 63.

4. Eastern Book Company v DB Modak (2008) 1 SCC 1.

5. ANI Media Pvt Ltd v Open AI Inc & Anr CS(COMM) 1028/2024 (Delhi HC, order dated 19 November 2024).

6. World Intellectual Property Organization, ‘Conversation on Intellectual Property Policy and Artificial Intelligence’ (WIPO/IP/AI/2/GE/20/1, 2020) https://www.wipo.int/en/web/frontier-technologies/artificial-intelligence/index accessed 26 June 2026.

7. SS Rana & Co, ‘Authorship vs Ownership of Work Under the Copyright Act: What is the Difference?’ (Mondaq, 2025) https://www.mondaq.com/india/copyright/1588878 accessed 26 June 2026.

8. Lexology / Global Practice Guides, ‘Indian Copyright Laws’ https://www.lexology.com/library/detail.aspx?g=e963324a-4b62-49ab-bd90-aeddb1aee2b0 accessed 26 June 2026.