Society has an obligation to ensure necessary research for which there is no profit motive, or for which the profit motive acts to distort the outcome of the research. Note that I did not say an obligation to fund such research, although that will frequently be the only way to make sure it happens.
In some cases, of course, such as research into MS or cancer, government needn't foot the entire bill, since private donations are significant, but for basic science where it isn't yet known whether any practical application might result, and especially for the investigation of potential collateral effects of developments which have moved into the realm of privately funded applied research, the onus is upon government to take the initiative, in the latter case because companies which stand to profit from the deployment of novel products cannot be trusted to investigate whether that deployment should be allowed.
Take, for instance, the effects of neonicotinoids on bees. How much can you trust studies saying that such concerns are overblown, when those studies have been paid for by the companies which profit from the sale of neonicotinoid-based products.
On the other hand, if the burden for such investigations into collateral effects were to fall entirely on government, that would leave no incentive for the companies involved to constrain their submissions to a likely few. Continuing with the chemical example, they might submit every compound they were able to synthesize and bog down the system to the point that there weren't enough qualified people on the planet to conduct the needed research. So at least some of that cost should be borne by those who stand to profit.
Moreover, it makes sense that the entire cost of determining whether a product should be allowed on the market should be paid for out of profits rather than taxes, so perhaps government should conduct whatever research is needed and then bill the company submitting the provisional product. To take this one step further, perhaps the payment of that bill should take the form of company stock, at market value, so the more often a company proposed a new product, necessitating a new round of research to assure the safety of that product, the more it would become publicly owned. In some cases, of course, the government could end up owning the company, since the cost of conducting the necessary research might exceed the company's market cap.
No doubt this would dramatically slow the introduction of new products to market, particularly in the chemical industry, and, at least in that case, that would be a welcome result. We need to be much more careful than we've been in the past with regard to what we introduce into our bodies and into the environment, taking the time to investigate collateral effects before allowing a new product onto the market.
Perhaps it is only in the chemical industry that such a process might be needed.