When it comes to protecting valuable ideas and products, traditional intellectual property (IP) systems (the legal protection designed to stop others from copying or profiting from an invention) have long been a roadblock for startups.
Founders have typically outsourced their IP management to external lawyers. But because of the high costs and slow processes, patents have often been viewed as something you just had to obtain, rather than a strategic advantage.
By combining traditional legal expertise with AI, startups are now able to protect their most valuable innovations without sacrificing speed, capital or control. This is what Lightbringer is working to do by pairing AI with expert patent attorneys.
“The old model is too slow, too opaque and too disconnected from how modern deeptech companies work,” says the company’s founder, Dominic Davies. The core failures are “speed and cost predictability,” he adds.
“What we decided to do is to build the sort of dream law firm that we would have liked to have when we were building a deeptech company,” he adds.
In an interview with Sifted, Davies and Ran Xiao, CEO and founder of AI spatial intelligence company Planarific, explore how using AI to secure patents is putting founders back in control of their IP strategy.
The traditional patent model changes with AI
When it comes to patent law, the legal industry has long relied on human processing time. For startups moving at a fast pace, this creates lengthy timelines.
“You can't have product cycles that are shortened to months or weeks using AI, while the patent process still moves at the traditional industry speed, which is three to six months,” Davies says.
“Engineers have finished what they’re working on and they've started thinking about something else,” he adds.
One of the main reasons for long traditional processes is billable hours, with patent lawyers often incentivised to charge for as many hours as they can. This “encourages them to be slow and to deliver the minimum while spending the most amount of time,” says Davies.
As a very small company, we don't have endless resources.
These costs were a big reason Planarific was discouraged from pursuing software patents through traditional lawyers. “It's costly to file and costly to defend,” says Xiao.
Planarific uses AI to convert aerial and ground imagery (such as drone photos) into 3D building models. The company helps local authorities, housing associations and property managers to scale up home retrofits and building analytics.
“As a very small company, we don't have endless resources, and we don't have specific people who can look at filing an IP and talk to patent lawyers,” Xiao adds. “That's a luxury we can't afford.”
In fact, many founders don’t know how much the traditional way of securing IP is going to cost until the invoice arrives.
Moving away from the billable hour to a flat fee model, the SaaS-based approach Lightbringer offers gives startups control over their budgets. The company charges a monthly or yearly fee per application.
“We can tell you exactly how much it's going to cost on a yearly basis, which is a game-changer for financial officers,” says Davies.
Using AI to reclaim IP
By using traditional patent lawyers, IP becomes something that is “set once and then thought about again in a year,” says Davies.
“The strategy needs to be a live document that gets updated every time you make a change to a product and every time you become aware of new competitor activity,” he adds.
Traditional lawyers also often lack the specialised background knowledge required when working with deeptech companies.
Using quantum computing as an example, Davies explains: “The inventor would approach the traditional attorney and convey their ideas. The attorney has to understand it with no help other than perhaps some quantum textbooks.”
AI changes this way of working, he adds. “The inventor comes, presents their technology to the AI, which understands it through its pre-trained data.”
Rather than selling an AI wrapper, which is a software application built on top of a third-party model such as ChatGPT, to traditional law firms, Lightbringer is built as an AI-native patent firm.
The real change here is ownership. We give the actual technical teams the ownership to manage what they're protecting.
It acts as a virtual IP department, using AI and proprietary data to do technical writing and research, while relying on human attorneys for final approval.
This means startups don’t drain internal resources when working on patents. “We chose to focus on the parts of our pipeline that are the most visible and that are the most likely to be copied or inspected by people who may want to do what we do,” says Xiao, who used Lightbringer to protect Planarific’s innovations.
“The real change here is ownership,” adds Davies. “We give the actual technical teams the ownership to manage what they're protecting.”
Inside the platform, technical teams work from a live view of their strategy, the competitive landscape and their invention pipeline, and can put direct questions to the AI about their own IP position.
Because the technology works from the underlying data rather than just the legal language around it, the applications that come out the other end are stronger, not just faster.
"The AI understands it and structures things correctly, and at that point the inventor and the attorney meet in the middle and check that both sides of the equation are done correctly,” says Davies. “That actually produces an even better quality outcome than was possible before AI.”
The value of patents in deeptech
Because large language models (LLMs) are able to write code and solve problems effectively, they are essentially “melting the moats” that traditional software companies previously relied on, says Davies.
It’s about showing your stakeholders and your potential investors that you've got something that's interesting and novel.
However, deeptech companies are able to secure their market position by protecting their innovations with IP.
The Planarific team looks at IP as more than “something you go to court and defend,” says Xiao. “It’s about showing your stakeholders and your potential investors that you've got something that's interesting and novel,” he says.
In sectors such as climate tech, the push for IP can sometimes conflict with what Xiao describes as a culture of open innovation.
Working alongside other startups, “people are quite open and they don't mind talking to you about what they're working on,” he says. “It's an area where if someone wins, everyone benefits. We're not really in the stage of fierce competition against each other.”
But securing IP remains essential for the long-term durability of deeptech companies, Xiao adds, particularly where R&D costs can be significant.
“You do have to have patents to make a viable business especially in sectors such as material science and manufacturing. That's just what you need for investment and for the business to run.”
