The Government's proposed rules for the conditional release of genetically modified organisms represent an astute balance, as much as anti-GM campaigners will continue to deny it. The rules need to protect individuals and the environment from the risks of GM research but not at the expense of erecting obstacles so formidable they stifle scientific work. They need, in other words, to create a framework to reinforce the Royal Commission on Genetic Modification's view that research should proceed cautiously. The potential for a civil lawsuit or a fine of up to $10 million for researchers who breach conditions for the release of GM organisms should be a sufficient deterrent to achieve just that.

Anti-GM groups will not buy that, of course. Green co-leader Jeanette Fitzsimons has described the liability regime as "weak and disappointing". Her party, and pressure groups that share its belief, will spare no effort in pushing their case to a select committee. There will be much teeth-gnashing and anguished appealing before the moratorium on GM releases is lifted in October.

In reality, however, anti-GM groups will accept nothing less than rules so draconian that research becomes impractical - and New Zealand deprives itself of potentially huge economic and social benefits by becoming a biotechnology backwater. They are still intent on undermining the royal commission's sensible guidelines, and even the importance of conditional releases for the gathering of information about new organisms. The select committee cannot let itself be hijacked to revisit such matters. Its job, one of the most important tasks of this term of Government, is to fine-tune the liability regime.

The Environment Minister has already conceded that a regime to cover all eventualities is not possible. Indeed, there is an element of "work in progress" about the Government's thinking. Wisely, however, it has rejected the case for special laws that would impose absolute liability for any harm that resulted from a GM release, whether that harm was proven or otherwise. The potential impact of a GM release mishap is, of course, more far-reaching than that of other environmental accidents. But if that is reason for a particularly strict civil liability and penalty regime, it does not demand patently draconian measures.

As it is, those who take legal action will not be required to prove negligence to be awarded compensation for a breach of conditions imposed by the Environmental Risk Management Authority. They will have to show only that those conditions have not been met. And even if the conditions were followed, standard liability regulations for environmental accidents must apply.

Jeanette Fitzsimons has also implied that the risk management authority should be held responsible if it approves a release that causes damage. That is an equally impractical concept. Anti-GM groups would best serve their cause by contributing sensibly as the authority considers, case by case, conditions to be applied to the release of GM organisms. These can include restricting the location and size of GM crops, not allowing flowering (so that contamination is avoided), importing only one sex of an animal (to avoid breeding), and ensuring that medical experiments are strictly monitored.

Researchers will propose their own rules for field trials. The response of the risk management authority, and the conditions it imposes, will, in the final analysis, determine the fate of biotechnology in this country. The strict liability regime will undoubtedly help to ensure that researchers act sensibly and safely.

Finally, New Zealand is on the right GM road.