Philips and DeepIP: Exploring AI tools in Patent Practice

About Philips

Royal Philips is a global health technology company focused on improving people's health and well-being through meaningful innovation. Protecting these innovations requires an intellectual property organization that combines deep IP expertise, technical understanding, and close collaboration with research, development, and business teams.

Within Philips' Intellectual Property & Standards organization, AI is being implemented as part of a broader digital transformation aimed at improving the quality, consistency, and effectiveness of IP work while maintaining rigorous professional and security standards.

This article reflects Philips' experience using DeepIP as one of several AI solutions supporting patent-related activities.

The Opportunity

At Philips, AI supports patent professionals throughout the patent lifecycle by rapidly analyzing large volumes of technical and patent-related information. It assists attorneys in understanding inventions, reviewing prior art, exploring claim strategies, preparing initial drafting material, identifying inconsistencies, and supporting patent prosecution activities.

By reducing the time spent on information gathering and routine drafting tasks, AI enables patent professionals to focus on what creates the greatest value: applying their IP expertise to develop stronger protection strategies, improve patent quality, and better protect Philips' innovations.

AI is an assistant, not a decision-maker. All outputs are reviewed and refined by qualified patent professionals, who remain fully responsible for the final work product.

Philips' Approach to AI

Philips has identified various opportunities to transform the IP lifecycle to improve quality, speed and impact.

Several principles guide the company's adoption of AI within Intellectual Property:

• Patent professionals remain responsible for every IP decision.
• AI-generated outputs are reviewed before being relied upon.
• Security, confidentiality, and governance are essential requirements.
• AI should enable patent professionals to focus on higher IP work.

These principles apply regardless of which AI technology is used.

‍DeepIP in Practice

As part of its broader AI strategy, Philips evaluated a range of AI solutions to support different activities across the patent lifecycle. DeepIP was selected for specific patent-related use cases because it aligned with Philips' functional, security, and governance requirements.

Today, DeepIP supports patent professionals in activities such as:

• understanding invention disclosures and other technical documents;
• reviewing and analyzing prior art;
• assisting in the preparation of Office Action responses; and
• generating initial patent application drafts.

DeepIP is one of several AI tools used within Philips' Intellectual Property organization. Philips continues to evaluate emerging AI technologies and expects different solutions to complement one another across the IP lifecycle as AI capabilities continue to evolve.

"AI is an assistant, not a decision-maker. All outputs are reviewed and refined by qualified patent professionals, who remain fully responsible for the final work product."

Current Use Cases

Understanding the Invention

A thorough understanding of an invention is the foundation of a strong patent application. DeepIP assists patent professionals in analyzing invention disclosures, technical documents, prior art, and supporting materials to identify key inventive concepts, technical advantages, and potential points of distinction over existing technology.

By providing structured summaries, highlighting important technical aspects, and identifying topics that may warrant further discussion with inventors, DeepIP enables patent professionals to build a deeper understanding of the invention at an earlier stage. This supports more focused inventor discussions, stronger claim strategies, and ultimately higher-quality patent applications.

Patent Drafting

Patent drafting remains a highly specialized activity. DeepIP is primarily used to generate initial material that patent professionals subsequently review, refine, and substantially develop.

DeepIP can assist with activities such as:

• identifying potential inventive concepts;
• generating initial claim structures;
• drafting supporting specification text; and
• assisting with dependent claims.

Patent professionals remain fully responsible for the final application, including claim scope, filing strategy, and technical accuracy.

Patent Prosecution

AI can also support preparation for patent prosecution by assisting patent professionals in reviewing Office Actions and organizing relevant information before interacting with the patent office or external patent counsel. This allows Philips patent professionals to provide sharper responses, clearer instructions, align internally on prosecution strategy, and have more focused interactions with outside counsel.

The assessment of Office Actions, amendment strategy, and communication with patent offices continue to rely on the judgment and expertise of qualified patent professionals.

Ultimately, AI is strengthening the role of the patent professional.

Observations to Date

Philips has observed several positive outcomes from integrating AI into selected patent workflows. While adoption continues to evolve, AI has already demonstrated value in supporting patent professionals throughout the innovation process.

Observed benefits include:

• enhancing the quality and consistency of work products;
• supporting more informed IP decisions;
• facilitating collaboration throughout the patent process; and
• reducing time spent on routine activities, enabling greater focus on strategic work.

Early experience indicates that AI can improve selected drafting and prosecution workflows. While benefits vary depending on the task and individual user, AI enables patent professionals to devote more attention to strategic work that benefits from their IP expertise and judgment.

Looking Ahead

Philips expects AI capabilities to continue evolving across the patent lifecycle, supporting activities such as invention evaluation, patent drafting, prosecution, portfolio analytics, knowledge management, and collaboration with external patent counsel. As AI technologies continue to mature, Philips will continue to evaluate how different solutions can best support its IP professionals and integrate into existing workflows.

At the same time, Philips believes that successful AI adoption depends on more than technology alone. Training, change management, effective governance, robust security, responsible use, and continuous development of AI capabilities within the IP profession will remain essential to realizing AI's full potential.

Ultimately, AI is strengthening the role of the patent professional. By combining AI with human judgment, technical understanding, and IP expertise, Philips aims to create higher-quality intellectual property, make better-informed IP decisions in the most efficient and effective way, and enable patent professionals to focus on the strategic work where they create the greatest value.

DeepIP was selected for specific patent-related use cases because it aligned with Philips' functional, security, and governance requirements.