With innovation at the core of what venture-backed companies are building, it’s natural that vetting of intellectual property should hold a prominent place in a fund manager’s due diligence process.

For the early-stage companies that H/L Ventures backs, founding partner Oliver Libby wants to see whether they have a plan for how to document and protect their IP assets. “A lot of the companies we’re looking at are so early that they haven’t had the money or time to properly [document their] intellectual property,” he says. “That’s OK and not uncommon, depending on the industry.”

If a founder doesn’t already have a great law firm to manage that process, H/L can bring one to the table. “We can’t move forward with an investment if we are not confident that there is either documentation of the intellectual property or that there’s a credible plan to do so.”

For Libby, IP is part of “a defense in depth” rather than a silver bullet. That can consist of multiple defensive tactics, including having patents in process, documenting trade secrets internally and safeguarding internal cybersecurity. “All of that together should ladder up to good defense.”

Ensuring that employees have signed non-compete, non-disclosure and assignment discovery agreements is part of that process. Many early-stage start-ups either forget to secure or decide not to pursue such pacts because they think it’s too expensive, onerous or time-consuming, he notes.

The last thing a company that is dealing with advanced technology wants is for a disgruntled employee to leave and later claim to have contributed to the thinking behind a product and challenge the company’s ownership rights, says Libby.

The pre-IPO filings of a lot of large companies that went public with significant technology often include “one or two random little claims from early in the company from someone who said, ‘I was there for six months and a lot of the underpinnings of the key thing are my thinking.’ And it’s problematic if they don’t have an assignment agreement,” Libby explains.

For first-time entrepreneurs, the extent of their legal strategy might be “Uncle Jim who’s doing the legal documents and is a real estate lawyer,” he adds. “That’s almost always where you find the initial set of weaknesses in that process. And that’s OK. You can bring in the right people on top of that, but you have to understand where they are [in their process].”

Open-source headaches

Generative AI, a hot new area of deep technology, has recently posed IP challenges. In November, a class-action complaint was filed in the US District Court for the Northern District of California on behalf of open-source programmers against GitHub Copilot, its parent, Microsoft, and its AI technology partner, OpenAI, alleging violations of open-source licenses that require attribution.

Other companies besides Github Copilot are generating code using AI that’s been trained on existing code, which is someone else’s intellectual property. Inadvertent violations can be guarded against using tools similar to those universities have used to detect plagiarism in students’ work, says Vivek Ladsariya, a partner at SineWave Ventures. Most such companies, however, don’t use AI-generated code on its own, Ladsariya notes. “They’ll generate code and make myriad adjustments before deploying it.”

Using generated code on its own can create lots of vulnerabilities, or at least many bugs, because it isn’t properly integrated with the rest of the code base, he adds.

An entire subsegment of “explainability companies” such as Arthur AI, Fiddler and TruEra are built on demand for more transparent and trusted AI models that inspire greater confidence in model outcomes, says Ladsariya. He hastens to explain that open source is a relevant and valuable business model that well-regarded companies like ­DataBricks have built their products on.

“We were invested in DataBricks,” he says. “Despite the fact that the IP wasn’t owned by DataBricks, it was a good investment because the people running DataBricks really understood the IP. So they could build value-added layers on top that enterprise [customers] would care about.”

Innovation is often a highly iterative process that involves a lot of shared work with open sourcing, agrees Libby, who, in addition to heading H/L is a venture partner at SineWave. With open source technology prevalent in AI and other advanced deep technology tools, every backable company isn’t required to have its own completely new invention, he says. “Sometimes it’s how the product was built. It’s how they wielded that technology, how they threaded it together. And that can become protectable as well – like trade secrets or use patents.”

Like many VC firms, SineWave relies on an external law firm for its IP due diligence. Much of that entails reviewing not only employee contracts to confirm IP has been assigned to a company but also commercial contracts.

“In an enterprise sale, the contract is large and complex and you might inadvertently assign all of the IP to a customer,” Ladsariya warns.

IP due diligence for investments in products that emerge from university research, typically by a faculty member, present other complications.

“We can’t move forward with an investment if we are not confident that there is either documentation of the intellectual property or that there’s a credible plan to do so”
Oliver Libby
H/L Ventures

Under the terms of faculty employment, 95 percent of the time, therapies developed in this way are owned by the university, says Barry Burgdorf, a partner in Hogan Lovells’ life sciences practice. Once a faculty member discloses what he’s invented to the university’s technology transfer office, that office decides whether or not to pursue patent protection. While he’s not an IP attorney, Burgdorf works closely with IP lawyers in his firm.

“The faculty member has an obligation to assign the patent and all the related technology to the university,” he says. After filing for patents, the university will license a technology for commercialization, either to the faculty member who has launched a company or to an existing biotech firm that thinks the technology will work with something it’s already developing.

One of Burgdorf’s biggest clients is the Cancer Focus Fund, which, in partnership with the University of Texas MD Anderson Cancer Center, invests in new oncology therapies nearing their FDA phase 1 or phase 2 clinical trials.

“To successfully get a cancer therapeutic to market, you need more than one patent,” he says. “You’ve got a whole area, and you’ve probably assembled that through licensing from a university at the outset. Then maybe you’ve supplemented that initial intellectual property with IP from other biotechs, other enabling technologies or drug delivery technologies, depending on what your target is.”

After confirming the target of a clinical trial, Burgdorf looks at the patent portfolio a company has assembled to determine if it matches the clinical trial and will be able to get a given drug to market. He ensures a conflict-of-interests management plan is in place to preclude any “cloud on the actions of the faculty member and hopefully no dispute between [them and] the university.” Burgdorf also confirms there are no prior claims on the technology, such as by a governmental funding source.

In addition, Burgdorf scans the competitive landscape for prior art that might raise concerns. For a company preparing an immunotherapy for clinical trial, for example, he checks whether there have been announcements of similar treatments to confirm there’s a clear path to commercialize any technology the company has identified. “If we see any missing pieces of the puzzle, that’s obviously a red flag,” he notes.

Thorough legal diligence enables discovery of all the intellectual property that a company has properly documented. “A lot of this is possible because we’re not doing a hundred deals a year,” says Ladsariya of SineWave. He cites the breakneck pace at which giant funds would deploy $1 billion in a year and then raise another $2 billion and deploy that in a year. “Any time you’re behaving in that way, it’s impossible to be diligent in your process.”