To do so, they usually convene a panel of scientists to review published and unpublished safety data. Of 562 publicly available GRAS determinations voluntarily submitted to the FDA since 1998, the center analysis found that companies used such panels two-thirds of the time.
These panels, typically composed of three members, are meant to represent the scientific community at large. And they have great power because their decision is the final word on the use of the ingredient. Once the panel deems a new additive as GRAS, it can go into an array of foods that end up on supermarket shelves, with no notice to or review by the FDA.
Critics say that gives food companies an incentive to turn to experts they believe will look kindly upon their ingredients and gives scientists incentive to do so.
"If I know that my paycheck is coming from a specific source, and I've been doing that for years and years, and that is what feeds me and my family, it becomes really difficult for me to be totally independent of the hand that is feeding me," says Erik Olson, senior strategic director for health and food at the Natural Resources Defense Council.
Many scientific consultants dispute accusations that they are conflicted, arguing instead that they are the most qualified and most experienced scientists for the job.
"If you're good at something, of course you're going to be in demand," says Joseph Borzelleca, who has appeared on 41 percent of the panels convened in the past 17 years — three times as many as the next most hired expert.
Borzelleca and other scientists say their reviews are thorough and that the GRAS process ensures the safety of food additives.
"As long as you adhere to science-based review," says John Thomas, a scientific consultant, "then I don't think there's a better peer-reviewed process in place."
The world of GRAS panelists is a small one. A Center for Public Integrity analysis found that the top 10 most frequently hired panelists have each sat on two dozen or more panels. Oftentimes, the same team of experts serves on panels together.
"There's a reason you keep going back to the same people, and that's because these are people with sterling reputations, impeccable credentials and they know what they're doing," says James Heimbach, who has convened GRAS panels for decades.
But "the fact that there's ... repetition and there's familiarity ... could potentially breed a conflict," adds Steve Morris, acting director of natural resources and environment for the Government Accountability Office, which published a report in 2010 that cited financial conflicts of interest in the GRAS system as a concern.
The center's analysis very likely captures only a fraction of all expert panels convened to establish the GRAS status of additives. That's because companies can make safety evaluations in secret without ever telling the FDA.
Companies are allowed to hire a single consultant to sign off on safety determinations or rely on the judgment of their own experts — and did so about a third of the time.
The center identified at least 10 GRAS panelists — including four of the most frequently hired experts — who have in the past had ties to the tobacco industry, according to a review of tobacco industry documents archived by the University of California, San Francisco. Two of the 10 scientists were once full-time employees of big tobacco companies.
Some of these panelists said that their work for the tobacco industry was limited to evaluating the safety of cigarette additives or newly developed cigarette products that tobacco companies thought would be less dangerous. They stressed that they did not defend the safety of cigarettes in general.
Borzelleca's work for the tobacco industry dates at least to the early 1980s.
An RJ Reynolds memo from 1984 notes that Borzelleca "has been secured by the tobacco industry to represent our position" during discussions with the Department of Health and Human Services about cigarette additives.
Two years later, the company envisioned Borzelleca as its "main spokesman" if a list of cigarette additives submitted to HHS were to be leaked to the press and "there is sustained and intense media coverage of ingredients issues," according to a confidential memo regarding the tobacco industry's "public relations strategy."
"My advice to the tobacco industry was that the GRAS substances they were using are safe when ingested," Borzelleca wrote in an emailed response to center questions, "but I could not comment on their effects when they were subjected to the high temperatures of a lighted cigarette, a position that I still have."
For the full investigation, head to the Center for Public Integrity's site. Follow the center on Twitter @Publici or sign up for its newsletter.