At its annual Tech & Sourcing Summit in New York on 4 November, a technology transactions and commercial contracts team brought in-house lawyers and sourcing specialists together to consider how artificial intelligence is changing outsourcing, vendor relationships and technology agreements. The programme ranged from legal ethics and cybersecurity to quantum computing, intellectual property and the risks of becoming tied to a supplier’s technology.
A central theme was the ...
Continue Reading This Article
Enjoy this article as well as all of our content, including reports, news, tips and more.
By registering or signing into your SRM Today account, you agree to SRM Today's Terms of Use and consent to the processing of your personal information as described in our Privacy Policy.
That shift complicates contracting. The risks of a generative chatbot, an automated workflow and a more autonomous system are not necessarily the same, so protections need to reflect the particular use and its consequences. Contracts may need to set out how a system works, how its performance will be tested, who is accountable for decisions and failures, and what remedies apply when agreed standards are not met.
The National Institute of Standards and Technology’s voluntary framework for generative AI supports this emphasis on controls tailored to context. It recommends managing risk across a system’s lifecycle, from defining responsibilities and testing before deployment to monitoring performance and responding to incidents. Its guidance also covers unreliable outputs, privacy and security, intellectual property, and risks from external models and data. It is not a binding regulation or a contract template, but offers organisations a structure for assessing systems and their supply chains.
Legal teams face a related challenge in their own use of AI. The American Bar Association’s Formal Opinion 512 says existing duties, including competence, confidentiality, communication with clients and supervision, continue to apply. Lawyers need to understand the tools they use, verify outputs and exercise their own judgement. Before entering client information, they should also establish how a system handles and retains data. AI-generated meeting records and summaries bring further questions about disclosure, consent, accuracy and the protection of confidential communications.
Outsourcing arrangements are changing alongside these internal practices. When a supplier uses AI to deliver services, customers and providers may consider consumption-based charges, fees tied to outcomes or gainsharing linked to productivity. But pricing is only part of the negotiation: parties also need clarity about when AI is used, what human oversight is provided, how performance can be audited and who bears responsibility for the service.
The summit also examined quantum computing, a developing field that still faces substantial technical hurdles but could eventually threaten some encryption methods. The US National Institute of Standards and Technology warns that the arrival of a computer capable of breaking widely used cryptography remains uncertain, while stressing that migration can take years. Encrypted information stolen today could be retained for possible decryption in the future. NIST finalised its first three post-quantum cryptography standards in 2024 and advises organisations to begin identifying affected systems and working with technology suppliers on preparations.
Intellectual property is another unsettled area. The US Copyright Office’s 2025 report says protection for work produced with generative AI depends on sufficient human authorship: a person’s creative contribution must shape enough of the expression. Prompts alone are not sufficient, while human-authored material or creative modification within an AI-assisted work may qualify. The distinction makes it important for agreements to specify rights in inputs, outputs, models, training data and improvements, and to address whether customer information may be used to train or refine systems.
Dependence on a supplier can persist even when a customer owns an AI solution, if its operation relies on a provider-controlled model, platform or integration. UK Government Digital Service guidance on cloud lock-in advises organisations to assess the benefits of provider-specific services against the cost and difficulty of moving elsewhere. It recommends considering dependencies and exit plans during design and procurement, rather than waiting until a contract is ending. Data access, portability, continuity and transition assistance therefore need both contractual provisions and practical arrangements behind them.
Across the summit’s discussions, the common message was that AI adoption brings operational and legal decisions that cannot be left to technology teams or suppliers alone. Organisations will need to define expected results, maintain oversight and plan for change as systems, threats and commercial relationships evolve.
Source: Noah Wire Services



