Breaking Down Paywalls to Accelerate Research Powered by AI

Analyzing data

Courtesy of Getty Images

As 2023 rolls on, measures approved by the Biden-Harris administration are now at the proverbial rubber-meet-road phase as agencies scramble to figure out how to implement them.

Taxpayer-funded research in biosciences will no longer sit behind paywalls. Instead, all of this data will be publicly available for no cost. Agencies from the NIH to NASA must implement the updated guidance by December 31, 2025. Though agencies have a few years to officially transition, the work has begun.

Though the policy may be a shift at the federal level, big pharma won’t feel much of a blip, said one expert in the world of biopharma publishing.

Vikram Savkar, senior vice president and general manager of the Medicine Segment of the Health Learning, Research & Practice business at Wolters Kluwer, where he represents and manages research tools in biopharma, like the New England Journal of Medicine, told BioSpace it’s important to contextualize what the policy will actually do.

“All of us who have a stake in the space care about publishing but mostly impact on patient outcomes,” he said. “Anything that accelerates translation, from published research to drug development to clinical pathways in hospitals, is positive.”

Savkar went on to explain the subtly in what makes acceleration meaningful, careful to parse out that acceleration, in and of itself, isn’t a universal good in this instance.

For big-budget pharmas and institutes, paywalls haven’t ever been a barrier to accessing research; thus, though this policy will tear those walls down, the effect will be minimal.

Instead, the meaningful acceleration will come in text and data mining, much of it powered by AI, he said.

“Artificial intelligence applied to large volumes of peer-reviewed literature, drawing automated conclusions, tools applied to large bodies of content … That I think is where you will see acceleration because the policy will require all of that content to be made available in the same format in the same public databases,” he said. “It will require the data sources within those articles to be available. I think this is where the interesting transformation is going to be. And we'll see some very promising new developments for researchers and for us as providers as well.”

Though some opponents voiced concerns the U.S. would effectively be giving away intellectual property to the rest of the world, Alondra Nelson, Ph.D., who was OSTP director when she issued the policy, countered. 

“When research is widely available to other researchers and the public, it can save lives, provide policymakers with the tools to make critical decisions, and drive more equitable outcomes across every sector of society,” she said in August 2022. “The American people fund tens of billions of dollars of cutting-edge research annually. There should be no delay or barrier between the American public and the returns on their investments in research.”

Ultimately, her memo issued three broad recommendations for all federal agencies, including the NIH and FDA, such as:

  • Update public access policies as soon as possible, and no later than the end of 2025, to make publications and their supporting data resulting from federally funded research publicly accessible without an embargo on their free and public release;
  • Establish transparent procedures that ensure scientific and research integrity is maintained in public access policies;
  • Coordinate with OSTP to ensure equitable delivery of federally funded research results and data.

While these concerns about implementation may play out their relative validity as time goes on and agencies work out the details, Savkar said his primary concerns in rolling out the policy lie in managing expectations.

“We don't want false expectations leading to sudden disappointment,” he said.

In essence, patients shouldn’t expect to see algorithms come up with new drugs to market immediately. Instead, meaningful acceleration paired with powerful AI tools in existence and yet to be developed concomitantly with the policy will produce results … in time.

“We are ultimately talking about matters of life and death, drug development, clinical pathways that can make a profound and transformative difference in people's lives, life outcomes. It's not the sort of situation where you would want to throw a ChatGPT at it and say, ‘OK, well let's do what it says.’”

Making content, even once federal agencies have standardized the formats by which data will be available (December 2025), will not suddenly produce its own translation, he said.

“We all have to fasten our seatbelts and be in for a proper journey to develop the technologies that will take advantage of this content,” he said. “If we do that, in a rigorous and robust way, a way that respects what has always distinguished this space, which is editorial rigor, peer review, distinguishing quality for disinformation … I think at the end of a couple of years, we can find ourselves in a very productive place.”

Back to news