AI has become a defining force that is reshaping intellectual property (IP) law across patent prosecution, copyright protection, trademark strategy, and trade secret management. For patent attorneys, AI presents a dual challenge: it is simultaneously a subject of legal controversy and an increasingly indispensable professional tool.
Between 2014 and 2023, approximately 54,000 generative AI-related patent families were filed worldwide, and AI-related patent activity continues to accelerate. Among IP professionals, AI tool adoption surged from 57% in 2023 to 85% in 2025, marking a decisive shift from experimentation to embedded workflow integration.
The inventorship question
The US District Court for the District of Columbia affirmed in Thaler v Perlmutter that human authorship is a foundational requirement. Courts and patent offices in the US, the UK, the EU, and Australia have consistently rejected applications claiming AI as the inventor, affirming that current statutes require a natural person as the named inventor.
The legal standard has not shifted, but its practical burden has grown. When AI is central in R&D, companies must document which humans define the problem, select parameters, and make the final inventive decisions. For patent attorneys, this creates both a compliance obligation and a counselling opportunity: clients who integrate AI into product development must be advised to maintain time-stamped records, lab notebooks, prompt logs, and design rationales, to demonstrate where human judgement shaped the invention.
Copyright, training data, and infringement risk
The more commercially significant controversy involves the AI training process. Many generative models were trained on datasets scraped from the internet, often including copyrighted works without the rights holder’s consent. In June 2025, Disney Enterprises and Universal filed suit against Midjourney, alleging its model was trained on copyrighted characters and images without authorisation and arguing that the AI engine was built on systematic infringement.
WIPO has acknowledged that the rise of generative AI is accelerating the need for stronger copyright infrastructure to protect creators while permitting innovation to flourish.
AI as a practice tool
AI-driven tools now allow patent attorneys to:
Automate prior art searches across global databases in multiple languages;
Conduct patent landscape analyses;
Generate predictive litigation analytics; and
Flag inconsistencies in claim drafts.
Portfolio management platforms centralise patent data, surface renewal deadlines, and continuously monitor competitor filing trends. In transactional work, AI accelerates IP due diligence by reviewing applicable legal frameworks and identifying potential infringement exposure to reduce client costs.
These capabilities amplify attorney effectiveness but do not replace professional judgement. The interpretive and strategic work of IP practice remains firmly in human hands.
Governance, trade secrets, and the road ahead
AI adoption does not occur in a regulatory vacuum. The EU AI Act imposes obligations around transparency, training data quality, and organisational AI literacy: compliance demands that are becoming a subject of IP legal counsel. Trade secret protection presents related concerns; without strict contractual controls, confidential information transmitted to third-party AI vendors may lose its protected status. Strong non-disclosure agreements, explicit prohibitions on vendor training using client data, and audits of employee AI tool usage are becoming standard elements of sound IP practice.
AI’s absence from IP practice increasingly requires justification rather than its presence. Patent attorneys who invest in understanding the legal implications of AI-generated innovation and the strategic capabilities of AI as a professional tool will be better positioned to serve sophisticated clients in the future.
