The agreements that move technology between companies, from software and SaaS to IP and patent licensing, data and cloud, and AI and development deals, drafted so the rights, the risk, and the value end up where you intended.
A technology deal is a sale of something nobody can inspect: code, data or a model. The contract is the only place the asset exists in a form the parties can test, so its words carry the weight a survey or a title search carries elsewhere. If the vendor fails, Section 365(n) of the Bankruptcy Code lets a licensee keep using licensed patents and copyrights so long as it keeps paying, but the estate owes it no support or updates.
The counterparty's form is written for the counterparty. A service credit described as the sole and exclusive remedy for downtime is a price adjustment, not a remedy; without a right to terminate for chronic failure, the customer has agreed to accept a few percent back for an outage that stops its business. We read the form for sentences like that, then the grant, then the exit.
One partner holds the file from term sheet to signature and brings in the firm's privacy and M&A lawyers as the deal requires. The agreement has to work years after closing, on the day the relationship stops.
Software licenses, SaaS and cloud subscriptions, and service agreements on either side, with service levels, data ownership, security, liability caps, and termination drafted to match the commercial deal, whichever side's paper it starts on.
Inbound and outbound licensing of patents, copyrights, trademarks, and trade secrets, with scope, field of use, territory, royalties, and improvement rights drafted so the value you mean to keep is not licensed away.
Development, integration, professional-services, and joint-development agreements, defining deliverables, acceptance, ownership of work product, and rights in jointly created technology before the work begins.
Data licensing, hosting, and AI and machine-learning arrangements, addressing data ownership and use, privacy and security obligations, and rights in training data, models, and outputs, with our data-privacy and AI teams.
Collaboration, OEM, reseller, distribution, and platform agreements that combine the technology of two companies, allocating IP, revenue, exclusivity, and exit so the alliance survives a change in either party's strategy.
IP and technology diligence, representations, and transfer mechanics in acquisitions and financings, confirming the target owns or can use what it sells, run alongside our M&A team so the tech terms and deal terms align.
An employee's code belongs to the employer. A contractor's belongs to the contractor, because the statute limits commissioned works for hire to nine categories that do not include software, and even those need a signed agreement. The fix is a present-tense assignment, "hereby assigns", before the first commit.
Field of use, territory, term, exclusivity, and rights to improvements decide what a license really conveys. An over-broad grant can give away the value you meant to keep, and a narrow one can starve the deal.
A cap at twelve months of fees with a consequential-damages exclusion means the vendor's exposure for a breach is roughly what you paid it, and nothing for your customers' claims. Negotiation lives in the carve-outs: indemnities, confidentiality and data-security failures, often with a separate, higher cap for those.
The Copyright Office will not register a work without a human author, and the courts agree. A contract cannot grant copyright in purely generated output, only contractual rights between the parties. A data license should say whether training a model is a permitted use, because the parties will disagree.
Open-source components bring license conditions that can reach into proprietary code. Diligence on the license stack avoids surprises in a financing or acquisition, where unmanaged open-source use becomes a real exposure.
What happens on termination, breach, insolvency, or a change of control should be drafted at the outset. Source-code escrow, transition assistance, and data return matter most exactly when the relationship ends badly.
Before any drafting, we establish what there is to license: the assignment chain from every developer and contractor, the open-source components sitting inside the code, and which rights the company must keep to protect its own roadmap.
The grant clause is built term by term: exclusive or not, field of use, territory, sublicensing, rights to improvements and derivatives. Each word widens or narrows what leaves the company, so each is chosen rather than inherited from a form.
With the grant fixed, the fight moves to the terms that decide disputes: service levels and remedies, data ownership and security, indemnities and liability caps, and, in AI and data deals, who holds rights in training data, models, and outputs.
Termination, breach, insolvency, and change of control are drafted while the parties still agree, with source-code escrow, data return, transition assistance, and survival terms set so the agreement performs on the day the relationship stops.
In a technology deal, the contract is the asset, so we draft it like one.
A technology transaction is any deal whose value turns on rights to technology rather than on a physical asset: a software license, a SaaS subscription, an IP or patent license, a development or services agreement, a data or cloud arrangement, or a collaboration in which two companies combine technology. The defining feature is that the contract has to allocate ownership, usage rights, and risk in intangible assets, which is a different discipline from a conventional supply or sale agreement.
An assignment transfers ownership of the technology outright, so the assignee holds the asset and the assignor keeps nothing unless the agreement says otherwise. A license keeps ownership with the licensor and grants the other party defined rights to use the technology for a scope, a term, a territory, and a fee. Choosing between them, and drafting the scope precisely, is often the most consequential decision in the deal, because an over-broad license can give away the value the company meant to keep.
For a meaningful SaaS engagement, yes. The standard form a vendor offers is written for the vendor, and the terms that carry the most weight are rarely the ones anyone reads first: service levels, data ownership and security, limitation of liability, indemnities, and what happens on termination. We negotiate SaaS and cloud agreements on either side, mapping the commercial deal to terms that hold up when uptime slips, data is breached, or the relationship ends.
As four separate questions, each with its own clause. Who owns the data, including the derived and aggregated data the vendor would like to keep. What the vendor may do with it, with model training named rather than left inside the word use. What happens on a breach, where the GDPR gives a controller 72 hours to notify its regulator, so the processor's duty to notify the controller has to be measured in hours and written down. And who holds rights in the model and its outputs, remembering that purely machine-generated output has no copyright to assign. For a customer with EU operations, the Data Act adds a right to switch cloud providers, with switching charges banned from January 2027. We draft these with the firm's privacy lawyers, against the regime that actually applies to the data.
In a technology acquisition, the IP and technology are often the reason for the deal, so diligence and the transaction documents have to confirm that the target owns or has the right to use what it is selling. We review license stacks, open-source usage, assignment chains, and key contracts, and we draft the IP representations, warranties, and transfer mechanics so the buyer gets clean rights and the seller's exposure is defined. This work is run alongside the firm's mergers and acquisitions team so the technology terms and the deal terms stay aligned.
The recurring problem is imprecise scope: a grant that is broader or narrower than the parties intended, ambiguous rights to improvements and derivatives, unclear ownership of jointly developed work, and silence on what happens at termination or on a change of control. Each of these surfaces only when the relationship is under strain, which is exactly when the cost of the ambiguity is highest. We draft the scope and the exit terms deliberately at the outset so the agreement does its work when it is tested.
Open-source license compliance, policy, and diligence, managing the obligations that copyleft and permissive licenses impose on proprietary code before they surface in a deal.
Explore →Patent, trademark, copyright, and trade-secret strategy, plus the licensing and fair-use agreements that monetize and defend the assets carrying a company's enterprise value.
Explore →Counsel on AI development, deployment, and governance, covering rights in training data, models, and outputs, and the regulatory frameworks emerging around machine-learning systems.
Explore →Term sheets, funding rounds, and investment documentation for the technology companies whose IP and contracts a financing depends on, with securities compliance built in.
Explore →Reviewed by Jonathan D. Herpy, Sr., Senior Partner · Last updated October 2026 · Educational overview · Not legal advice
One partner is accountable for the whole matter, and you have that partner's direct line. Chicago, London, and Dubai work to the same standard.