A software development agreement is a contract between a customer and a software development company or independent programmer governing the creation of new software or the adaptation of an existing system. It specifies what will be developed, who is responsible for the work, when deliverables must be provided, how they will be tested and what rights each party will hold in the software and source code. Its purpose is not only to prevent delays and payment disputes, but also to ensure that the completed product can be used and further developed in accordance with the customer’s needs.
The distinction parties often overlook is that paying for development, owning the intellectual property rights and receiving a copy of the source code are three separate matters. A customer can pay a developer to complete a system without acquiring copyright ownership, and can acquire rights in software without receiving the passwords, documentation or source files needed to operate it. A properly drafted agreement should deal with each issue separately.
What Should a Software Development Agreement Cover?
A software development agreement may govern a website or mobile application, an enterprise system, an interface between systems, an algorithm or the customization of a technological product for a particular customer. The customer expects a business outcome, while the developer must translate that need into a technical specification and perform work that may involve uncertainty. The agreement should allocate responsibility for specifications, provision of information, decisions, testing of deliverables and changes required during development.
Commissioning a new software product should be distinguished from subscribing to an existing service. A development agreement focuses on the creation process, acceptance testing and rights in the resulting work. By contrast, a software-as-a-service (SaaS) or cloud services agreement focuses on continuous service delivery, availability, support and return of information. A project may include both, but the development obligations should be distinguished from the service terms that apply after launch.
Software development contracts combine contract law with Israeli computer and internet law, intellectual property, confidentiality and information security. A contract based solely on a technical quotation may therefore fail to resolve disputes over code ownership, components brought into the project by the developer or the customer’s ability to continue the project with another supplier.
Software Specifications and Defining the Development Scope
The specification document is the starting point for understanding the developer’s obligations. It should identify principal functions, user categories, systems to which the software will connect, authentication methods, performance requirements, operating environments, APIs, security requirements and data to be migrated from previous systems. A general description such as “customer management system” does not explain what will constitute completion or what falls outside the project’s scope.
Responsibility for preparing specifications should reflect the actual working arrangements. In some projects, the customer provides a detailed specification and the developer undertakes to implement it. In others, the developer is engaged to undertake a paid specification phase and submit a document for approval before coding begins. Where the customer must cooperate by providing content, granting access or making design decisions, the deadlines and consequences of delay should be defined. A developer should not automatically bear responsibility for a delay caused by material that was not provided, but cannot simply claim that the customer “failed to cooperate” without documenting the request.
For complex projects, binding requirements should be distinguished from possible future features, deliverables should be specified for each stage and excluded services should be identified. If the customer also expects source code, documentation, screen designs, design files or cloud configuration to be delivered, those items should expressly form part of the deliverables rather than being assumed to accompany the functioning software.
Development Methodology, Timelines and Changes to Specifications
The agreement should be suited to the development methodology. Where the project follows a fixed specification, defined phases, deliverables and delivery dates can be established. Under an Agile approach, requirements may change during the work, making it essential to specify how tasks are approved, what counts as a completed deliverable in each development cycle, who may change priorities and how budget overruns are handled. Calling a project “Agile” does not remove the need for defined scope, approval procedures and progress tracking.
A common dispute is whether a particular request involves correcting something already promised or adding new work for an additional fee. A change-request procedure can reduce that uncertainty by requiring a description of the proposed change, an estimate of its effects on price and timetable, and written approval before implementation. Approval through email or a project management system may suffice if the agreement states who is authorized to give it and how records will be retained.
Target dates should be distinguished from binding delivery commitments, and delays caused by the developer from those resulting from specification changes or reliance on third parties. The agreement can provide a cure period, revised milestones or a termination right for material delay. Liquidated damages for delay should be assessed in light of their amount and their connection to losses that could have been anticipated when the agreement was signed; attaching a penalty to every minor deviation from the schedule may be inappropriate.
Fees, Milestones and Payment Dates
Two common pricing models are a fixed fee for a defined scope and time-and-materials billing based on hours and resources. With a fixed fee, it is particularly important to establish what is included and what requires a change order. An hourly model calls for work reports, approval procedures, a budget ceiling or advance notice of overruns. A project may combine both models, for example a fixed amount for specifications and initial implementation and hourly charges for further development.
It is advisable to link part of the fee to delivery of testable results while avoiding a mechanism that allows the customer to withhold payment indefinitely because of minor defects. Payments may be associated with completion of a development phase, successful acceptance tests or the end of an initial rectification period. The parties should know whether payment constitutes approval of the deliverable and what happens if the customer fails to perform acceptance testing within the agreed period.
A separate issue is the relationship between payment and the transfer of intellectual property rights. The parties may agree that assignment takes effect only after full payment for the relevant deliverables, or establish phased usage rights during the project. Either way, the customer should not be left having paid substantial sums but unable to use partially completed deliverables if the project ends before completion.
Software Delivery and Acceptance Testing
Completing the code is not the same as delivering software that the customer must accept. A development agreement should identify what is submitted for testing, which tests will be conducted, who prepares test data and the time allowed for the customer to notify defects. Acceptance tests should be linked to the specification, the agreed uses and the severity of faults.
A material defect preventing an essential use of the system should be distinguished from an issue that can be corrected after delivery without delaying operation. A balanced agreement permits the developer to fix defects and resubmit the software, while addressing what happens if repeated attempts fail. It may include deemed acceptance after a defined period, provided that the customer has received a testable deliverable and a genuine opportunity to examine it.
The relationship between acceptance and the developer’s subsequent warranty obligations must also be defined. Accepting a development phase should not automatically release the developer from responsibility for a latent defect discovered later, but it does not mean that new features must be developed without payment. The contract should distinguish corrections included in the warranty period from separate maintenance or development work.
Intellectual Property and Rights in Development Deliverables
Allocating intellectual property rights is among the most important decisions in a software development agreement. The parties should identify which deliverables are covered by the rights being transferred: source and object code, specially developed components, specifications, documentation, designs, interfaces, test scripts and other materials, insofar as protected rights subsist in them. Ideas, methods and information do not all enjoy the same legal protection. The phrase “all intellectual property” does not, by itself, necessarily settle the parties’ rights.
Who Is the First Owner of Copyright?
Under Israel’s Copyright Law, 5768-2007 (English text), a computer program is included within the definition of a literary work, and copyright may protect its original expression, including its code. Section 34 provides that the employer is the first owner of copyright in a work made by an employee in the course of employment and during the period of employment, unless otherwise agreed. Under Section 35, by contrast, the first copyright owner of a commissioned work is generally the author unless the commissioning party and the author have agreed otherwise, expressly or impliedly.
There is therefore a difference between a company employing a programmer and one commissioning development from an independent contractor. Payment to a contractor does not, in itself, ensure that copyright has been assigned to the customer. When the work is commissioned from a software company, it is important to verify that the company has itself lawfully acquired the necessary rights from its employees and subcontractors so that it can transfer the rights it promises the customer. The legal protection of software code is discussed further in our guide to software copyright in Israel.
Background IP Versus IP Created During the Project
A central contractual distinction is between background intellectual property (Background IP), brought to the engagement by either party, and foreground intellectual property (Foreground IP), created specifically during the project. A developer may use previously developed libraries, software engines, templates or tools for numerous customers. A blanket assignment of ownership in every system component could conflict with existing rights or prevent the developer from continuing to use its own tools.
Conversely, where the customer funded bespoke development, the agreement should state whether ownership of that work passes to the customer and whether the developer may reuse the deliverable itself or parts of it. Rights in improvements to existing code, jointly developed components and interfaces connecting the new software to existing systems should also be addressed. A schedule identifying material background components can prevent disputes after the system is already in use.
Assignment, Licensing and Commercialization of Software
Full ownership is not the only solution. In some transactions, the developer may retain ownership of general-purpose code while giving the customer a broad licence, whether perpetual or time-limited, according to the agreed commercial terms. The licence should specify whether it is exclusive and whether the customer may modify the software, distribute it, integrate it into another product, permit affiliates to use it, grant sublicences or transfer the rights to a purchaser of its business. Without clear provisions, the ability to commercialize the software may be restricted even where it functions as intended.
Section 37 of the Israeli Copyright Law permits assignments and the grant of different types of licences. Under Section 37(c), a contract assigning copyright or granting an exclusive licence requires a written document. The agreement should identify the rights being transferred, when the transfer takes effect, whether it depends on payment and what each party may do after the transfer. Copyright ownership does not itself guarantee actual delivery of the source code and infrastructure needed to operate the software.
Patents, Trade Secrets and Other Rights
Some development projects may generate technological know-how, processes or inventions whose eligibility for patent protection should be considered. Not every software component or business idea is patentable. If inventions may arise, the agreement should allocate rights to file patent applications, responsibility for costs and duties to cooperate by providing information and signing documents. Where the inventor is an employee, the service-invention provisions of Israel’s Patents Law, 5727-1967, should also be examined; those provisions should not be assumed to apply in the same manner to an independent contractor.
Algorithms, methods, training data, specifications and engineering know-how may, depending on the circumstances, also be protected by trade secret law and contractual confidentiality obligations. Protection under Israel’s Commercial Torts Law, 5759-1999, requires, among other things, that the information not be public and that reasonable measures be taken to preserve its secrecy. For the distinction between employees, contractors and commissioning parties, see our guide to ownership of trade secrets in Israel.
Section 45 of the Copyright Law should also be kept in mind: moral rights do not apply to computer programs themselves. A development project may nevertheless generate separate graphics, texts or other creative works to which different considerations apply. The deliverables should therefore be examined by type rather than subjected indiscriminately to a single legal mechanism.
Third-Party Components, Open Source and Licence Dependencies
Commissioned software is rarely developed in isolation. It may incorporate databases, open-source libraries, commercial components, APIs, interface kits or third-party systems. The parties should identify which components are essential to operating the deliverable, which licences govern them and whether the customer will remain entitled to use them when the development relationship ends.
Using open-source code is not prohibited, nor does it automatically require the entire source code of the product to be disclosed to the public. Obligations depend on the applicable licence and how the component is integrated and distributed. Nevertheless, selecting a component without reviewing its licence may restrict commercial distribution, require attribution or trigger obligations to provide particular code. The developer should identify incorporated components, provide licence information and obtain prior approval before including components whose obligations are material.
A third-party licence may also be granted only to the developer or depend on an account under its control. Rights granted to the customer in contractor-developed software are of limited value if the system cannot continue using an essential payment component, mapping service, messaging service or software library. Wherever possible, essential accounts and licences should be held in the customer’s name or a clear replacement mechanism should be agreed.
Source Code Delivery, Documentation and Access Credentials
Software delivery does not end with providing a working environment or operating version. The agreement should state whether source code will be maintained in a repository owned by the customer throughout development or transferred at specified times, and whether delivery includes version history, build instructions, configuration files, architecture documents, deployment scripts and test suites. Code that cannot be built or installed without the developer’s exclusive knowledge may not allow the customer to continue development with another provider.
In addition to the code itself, access to Git repositories, cloud environments, servers, service accounts, domain names, monitoring tools, app stores and API keys should be addressed. A customer-owned account should be distinguished from a developer-owned account serving multiple clients. In some cases, a shared development environment with defined permissions will allow the customer to obtain the deliverables even if a dispute arises or the developer becomes unavailable.
Where the developer does not transfer source code on an ongoing basis, the parties may consider source code escrow, under which an up-to-date copy is deposited with a third party and released on agreed conditions. Escrow is no substitute for examining rights and licences and is of little assistance if the deposited code cannot be operated. In LiveDNS Ltd. v. Kobi (Sabach) Shahar, Civil Case 5150-05-14 (Tel Aviv–Jaffa District Court, 13 October 2020), a dispute over copied software code and its consequences for users of the software illustrated why the source of code and the rights attached to it must be verified rather than relying merely on possession of a copy.
Confidentiality, Information Security and Subcontractors
An external developer may be given access to existing code, business information, databases and the customer’s working environment. A confidentiality agreement should specify what information is confidential, who may access it, the permitted purposes of use and what happens to the material at the end of the engagement. The developer should remain able to use general professional knowledge and its own tools, but those must be distinguished from the customer’s confidential documents, code, models and data.
Where development involves access to personal information, the agreement should define the purposes of access, staff permissions, testing environments, security measures, use of real data, reporting of security incidents and return or deletion of information when the work ends. The development contract must be distinguished from the website’s privacy policy: the latter concerns disclosure to users, whereas the agreement specifies what the developer may do with information and on what terms. Where statutory rules governing disclosure to external processors apply, the parties must also consider the corresponding contractual and security duties. See our guide to information security and cyber incidents in Israel.
If the developer engages subcontractors, the contract should state whether this is permitted, whether advance approval is required and the conditions under which those subcontractors may access systems and information. Beyond confidentiality and security, the developer must ensure that it can also transfer to the customer rights in work created by subcontractors. A general statement that the developer is “responsible for all its personnel” may not resolve the need for assignments of rights and confidentiality undertakings from the people who actually carried out the development.
Defects, Warranties, Indemnification and Limitations of Liability
The agreement should define the warranty period after acceptance, the defects covered, response times and a documented method of reporting problems. A defect caused by failure to comply with the specification should be distinguished from one caused by a modification made by the customer, a third-party service or a new requirement. The developer should undertake to rectify failures to meet promised deliverables, without becoming responsible for every change in the system or its environment irrespective of its involvement.
Third-party intellectual property infringement raises a separate liability issue. If a component supplied by the developer infringes copyright or is subject to a licence that does not permit the intended use, the agreement should specify who will manage the claim, bear the expenses and provide the remedy, such as obtaining a suitable licence, replacing the component or modifying the software. Indemnity obligations should reflect the source of the risk rather than make the developer responsible for material supplied by the customer in violation of others’ rights.
Development contracts frequently contain liability caps, exclusions of lost profits or indirect losses, and duties to mitigate damage. The parties should assess how these provisions operate against the project’s material risks, particularly loss of code, breach of confidentiality, copyright infringement or exposure of personal information. A low cap may offer little effective recourse if a central deliverable is lost or cannot be used due to a third party’s rights. Conversely, unlimited liability for every malfunction may be disproportionate to the fees and the risks the developer can control.
Project Suspension, Termination and Transition to Another Developer
A development project may end before completion because of delay, a disagreement about scope, breach, business changes or the customer’s decision not to proceed. The agreement should distinguish termination for breach from termination for convenience, specify any cure period and explain how payment for completed work and delivered items will be calculated. It should address whether the customer may also receive partially completed deliverables that have not passed acceptance testing and the extent to which they may be used after termination.
Termination also requires a practical handover: source code and documentation, credentials and passwords transferred securely, return of equipment and files, removal of the developer’s access and proper treatment of personal and confidential information. Where necessary, a paid transition period with a replacement developer should be arranged, including technical explanations and cooperation in transferring the working environment. A contractual right to continue software development is of limited value if the material required to exercise it is unavailable.
A software development agreement works best when, on the signing date, the parties understand not only what happens if the project succeeds but also what happens if it stops halfway through. Defining deliverables, acceptance stages, intellectual property rights and source code delivery in advance enables the customer to obtain a usable product and the developer to understand both its obligations and the rights it retains.
International companies commissioning development in Israel, and foreign law firms seeking Israeli counsel for software development contracts, ownership disputes, source code handover or related litigation, may contact Drori Stav at office@dwo.co.il.
