In a recent webinar hosted by IPWatchDog, Tamar Gomez (Co-founder of Ankar), Boris Welzer (IP Director at Nexans, a multinational industrial firm), and Clint Mehall (Partner at Davidson, Davidson & Kappel, a US-based IP law firm) shared first-hand insights on the evolving role of AI in intellectual property workflows.
The discussion centred on one key theme: AI tools are essential for companies that want to stay competitive. However, with so many options available, it is not always clear what actually works in practice.
This blog highlights the most important takeaways from the conversation focused on two of Ankar’s software solutions, in particular:

• Patent drafting, which aims to streamline the creation of high-quality patent applications

• Infringement detection, which helps to identify potential patent infringements from publicly available information

Tailored onboarding is essential to long-term success.

Ankar's collaboration with Nexans, a publicly listed French industrial company, demonstrated the value of a highly tailored onboarding process. Nexans came in with a specific problem: identifying infringing products online using their existing patents. Rather than providing a generic dashboard or search tool, Ankar worked closely with the Nexans team to shape the system around their specific needs.

The result was a curated shortlist of only thirty matches, filtered for relevance and commercial importance. These results weren't just used by the IP team. They created value across departments — from design to marketing — becoming a shared source of intelligence for understanding competitors and identifying product blind spots.

The lesson is clear. To unlock the real value of AI in patent work, onboarding must go beyond account setup. It must involve deep integration with internal workflows, structured feedback, and hands-on iteration that results in a solution that works for each company.

AI Patent Drafting: solutions should adapt to the attorneys

True efficiency gains in AI-assisted patent drafting only emerge when the tool adapts to how attorneys already work. While AI can help teams draft patents faster, that speed is only valuable if it doesn't come at the cost of rewriting or reformatting the output.

As Clint Mehall pointed out, even within a single firm, attorneys have distinct preferences for tone, structure, and reference styles. If a tool ignores those nuances, it can slow things down rather than speed them up.

Ankar's patent drafting assistant addresses this directly. It allows teams to upload prior applications, reference patents, and templates so that outputs match internal expectations from the start. Users can also generate background sections, insert technical diagrams, define domain-specific terms, and tailor content to different jurisdictions — all within a structured, editable interface that mirrors their existing workflow.

When such alignments are in place, the time savings are substantial. Teams using Ankar have typically reported efficiency gains of up between 30-50 percent, with one user describing a return to traditional methods as “like going back to the Middle Ages.”

AI Infringement Detection: Solutions are only useful if results are actionable

Infringement detection is one of the most promising use cases for AI in IP, but it only creates value when the results lead to clear next steps. For Boris Welzer, a breakthrough moment came when Ankar's system surfaced product-level matches that were finally usable. After fifteen years of testing various tools, most had returned either vague overlaps or unmanageable lists that made follow-up nearly impossible.

Ankar's approach was different. By intelligently slicing claims and focusing search efforts, the platform produced a curated shortlist of potentially infringing products — each one backed by concrete context like images, company details, purchase links, and technical explanations. This level of clarity gave the Nexans team confidence to explore enforcement, turning detection into action.

Infringement tools only work if teams trust the outputs enough to use them. When done right, they can move IP from passive oversight into an active lever for protecting and commercializing innovation.

Security and trust cannot be an afterthought

Security and Trust in IP AI solutions was another important topic of discussion during the webinar. AI solutions operating in the IP and R&D space handle highly sensitive documents — from draft filings to proprietary data and strategy notes. Without strong controls and a trustworthy infrastructure, adoption can stall, no matter how effective the tools might be.

Ankar’s approach reflects a security-first design philosophy. Tamar Gomez emphasized how her experience at Palantir and Helsing informed this emphasis on compliance and trust. Ankar’s systems are built with full data segregation, enterprise-grade encryption, and strict access controls.

Ankar is certified under both ISO 27001 and SOC 2 Type II, giving enterprise clients confidence that their data is fully protected.

Overall Takeaway

This webinar underscored a shift that is already underway: AI is becoming embedded in the day-to-day operations of IP and R&D teams. But the teams that succeed will not be the ones who adopt tools fastest. They will be the ones who implement them thoughtfully, focusing on secure, tailored, and explainable systems that elevate rather than replace human expertise. Watch the full session here.