Managing U.S. legal risks - AI applications and new technologies in the space and satellite industry

Exciting new applications are occurring in numerous areas, including:

• space robotics, particularly for In-Space Servicing, Assembly, and Manufacturing (ISAM)

• avoiding collisions and monitoring of space debris

• satellites connecting in space (rendezvous, proximity operations, and docking, or RPOD)

• space exploration including moon and Mars rovers

• many analytics applications, as huge amounts of data are being gathered from space

Data sensed from space (from visual to radar-based) is a seemingly endless source of new data, and machine learning algorithms are being used to process satellite imagery, detecting and classifying Earth’s features for geographical information systems, classifying various land cover types in imagery, crop monitoring and predicting, wildlife conservation, disaster response, and in connection with many other use cases.

As with terrestrial AI, the technical challenges to creating and deploying space-based AI are considerable.  However, every year more progress is made at being able to control space activities both from the ground and in space itself, at collecting data in space that cannot be obtained from the ground and in furthering space exploration.  Having privately-owned, manned stations in space seems within reach, and productive trips to Mars or even asteroids could well happen in the next decade.

New types of risks

With these new applications, issues are arising that have not been encountered before, including the following:

• Liabilities for damage caused in space – this can arise in many ways from these new applications.  Satellite docking maneuvers are inherently risky without any ability to “see” where the satellites are going.  Collisions with space junk (aka, orbital debris) is an ever-increasing risk.  Damage will inevitably occur during space manufacturing, repair or in-orbit fueling, just as it does on earth.  Liabilities initially will mostly be limited to property damage, but as more humans go into space, injuries will occur, particularly in an environment that requires considerable protection of humans to avoid injuries that otherwise would occur.
• Risks based on dangerous materials, and how to assess the dangers – there will be dangerous or potentially dangerous materials being placed in space.  We have clients planning space-based depots filled with highly flammable fuel (“gas stations in space”) or using nuclear fuels to power batteries that can operate at low earth orbit (LEO).  Steps are being taken to reduce or manage these risks, and some of the risks (such as the nuclear battery) may be fairly low but agencies and counterparties may need to be convinced that is the case.
• Risks of reentry – Many space-based risks are managed by the simple approach of keeping the risk in space, where there are no people, no cities with lots of buildings that can be damages and limited personal property (largely just satellites, although they are quite expensive).  However, the increasing risks from space junk are leading to a trend of destroying that junk by having it burn up on re-entry.  But what if it does not burn up, and re-enters the atmosphere, making its way potential onto populated areas?  There have already been some incidents, and these will increase.
• Failures to complete or perform, and how it can be assessed or measured – It is generally pretty easy to measure performance on the ground.  Concepts such as delivery and acceptance testing work well since their occurrence can be measured and confirmed.  This is less easy in space, where there can be no objective way to resolve a disagreement between the parties for example whether a repair was completed successfully or whether a service such as conducting an experiment was properly performed. 
• Ownership-related risks – Disputes about ownership have occurred throughout all or recorded history, but usually there is law on what can be owned and who is the owner.  This is not the case in space.  The few treaties in effect just address the ownership of planets, and say nothing about ownership of materials detached from those planets, equipment constructed in space, mining camps that may be set up, etc.  The recent US-sponsored Artemis Accords do address more of these issues, but are only slowly obtaining international support and signatures.  Governments do not necessarily have rights to award concessions or licenses to key materials or locations, since the governments do not have anything like clean title. 
• Regulatory risks – There is a tapestry of international, national and regional regulations governing the launching of objects into outer space, spectrum and orbital resources, the licensing process for remote sensing and other space-related activities, as well as all of the other regulatory issues associated with the collection, processing and use of data generally.  National governments regulate private remote sensing of the territory within their national boundaries.  Failure to abide by such rules can result in civil penalties, among other risks.

Limitations of existing law – treaties are minimalist 

Very often risks are managed against the backdrop of existing law, which allocates risk in the absence of agreements to the contrary.  Sometimes agreements can reallocate the risk, sometimes the law over-rides, but in most cases there is some rule that applies, either directly or by analogy.  In space there is often no law that addresses many of the issues that are likely to arise.  The few international treaties that exist are at a very high level (the expression “fifty-thousand foot view” comes to mind, but of course the view here is much, much higher than that), and do not have much legal substance.  By US standards, where most laws together with implementing regulations are voluminous, in multi-volume sets, space law is minimalist at best - all of the applicable space treaties/laws fit in a single pamphlet, and there are questions as to what laws should apply and the liability regime governing certain activities. 

There are potential terrestrial laws that would apply, if it is clear which country’s law is applicable.  In many cases it is not clear, and even if the law of the US does apply it is not clear how the law should be different because of the different environment, circumstances and risks.  Some of the authors of this article teach a space law class in which we receive papers every year putting forth interesting ideas about what laws should apply and how to analogize from the laws governing the high seas, Antarctica or other minimally populated areas.  Many articles on these issues are readily accessible on the internet.  But the fact remains that the application of the law to space activities is very unclear, and it is also unclear how the various countries are going to go about establishing more law against which risks can be allocated.

In some key areas such as human spaceflight the US itself has intentionally decided not to have any substantive regulation until the end of the “learning period”, which may expire in October 2023 or may be extended.  And how many years will have to pass before precedents are established and accepted, to fill in the vast gaps that existing based on the limited law and regulation that does exist?  It is fairly well established that risks and costs are reduced as uncertainty is lessened by having more examples, more precedents, more data.  That will probably happen with respect to space, but for players seeking to allocate or manage risks today this seems too far off.

How to manage these risks – are contract provisions sufficient? 

Given the lack of legal precedents to rely upon, companies inevitably turn to contracts to allocate risks.  But without the backdrop of existing law that applies in most other contexts, the contracts have to be quite lengthy and inevitably have gaps – it is not possible in individual agreements to replicate the vast tapestry of law and experience that underlies some much of our current lives, commerce and relationships.  We have prepared several first of its kind (maybe the only one of its kind) agreements for space ventures.  The parties have struggled with the inevitable length and complexity of the documents, and also with the fact that so much is not explicitly addressed.   This approach at least can address the principal risks and the parties can agree to a dispute resolution process to deal with others. 

Our firm is slowing building up a library of agreements to deal with management of risk in these new environments, and we can at least offer clients documentation that addresses the new situations which they are encountering.

Will insurance be there to cover these risks?

Many risks in space are covered by insurance.  More than other areas, where one party or the other takes a particular risk, in dealing with space risks insurance is very often obtained, and the issues to be negotiated always include who has to pay for the insurance. 

However, now that we are moving outside the traditional areas for space insurance, mainly launch on rockets that have a launch history and risks of satellites being damaged or ceasing to operate in space, there is no insurance available for many of the risks.  They uninsured risks include risk to humans in space, damage from space debris, explosions, satellite collisions during RPOD, manufacturing operations not working, damage to space stations, the list goes on.  There may eventually be insurance for many of these risks, as there is more data and more experience. 

The authors have spoken with insurance companies and insurance brokers about insuring these new situations.  We can generally report that insurance companies are willing to step up to the challenge, but there isn’t much data to quantify the risks.  With limited data only certain risks can be insured, and the insurance is expensive due to the many uncertainties.  Insurance companies are expert at reducing data and risk assessments to policies with specific wording and identified costs, but without data and without risk assessments insurance can only be a partial solution. 

The demand is certainly there – there are many large companies operating in the space area who are not comfortable having every mission be a "bet the company” risk, and would certainly buy the insurance if it could be reduced to a policy with a price.  We suspect there will be more and more insurance over time (but many never be the equivalent of an “all risks” policy.)  Without the precedential data, though, at present in many situations there simply is no insurance to purchase.

New laws and regulations – will they help allocate risk?

At least in the US new laws and new regulations dealing with the space environment are slowly developing.  However, the new regulations are mainly in the area of licensing, which involves the granting of permission to engage in some activity, whether it is for launch or communications.  And even in the licensing arena, the regulators are dealing with requests for new and different types of licenses for which there is no precedent. Regulators generally grant licenses to operators or launch providers, the parties who traditionally had control over the operations and therefore could be given the responsibility for ensuring that appropriate steps could be taken to comply with the law or license requirements.  Increasingly, though, the traditional players are not necessarily the parties with control over the risks.  And many of the activities giving rise to risks are not regulated at the present time.

Where regulators are able to take action, they do use the ability to deny licenses to persuade companies to implement safety protocols, provide orbital debris plans and take other steps the regulators deem appropriate for space companies.   But aside from commercial launch laws and regulations, which directly address liability issues, most of the new regulation just deals with what activities are to be allowed, not the costs or damages that occur if something goes wrong.  And there are still many key areas for which there is no existing law or regulation, or even ideas about how to implement such law or regulation.

One of the areas of most significant risk, human spaceflight, is likely to be regulated once the “learning period” ends, whether in October 2023 or some extended date.  Until then, the law is based on informed consent – humans choosing to go to space take the risk, based on some actions taken by providers to make things safe as required by the limited regulations that do exist.  If substantive regulations are ultimately adopted, even if “opt out” is permitted, that will at least provide a basis for the parties to allocate responsibilities and risks.  That will reduce uncertainty, shorten contracts, improve the availability of insurance.  The question of course is whether the legislators or regulators know enough about this new environment to pass laws or regulations that are good ones, and that will help parties manage risk without stifling innovation or making going to space too risky or too expensive.

Next steps

If you work at a space company, or are planning to invest in one, we have assembled a lot of experience with new space situations, new areas in which risk needs to be allocated or assessed.  Our dedicated space team would be happy to share its experience and hard-won knowledge with you, and guide you in your upcoming adventures in the space arena.  We have commercial lawyers experienced with space contracts of many types, regulatory lawyers familiar with the licensing and rulemakings going on at the FCC, NOAA, FAA and other agencies across the globe, lawyers who advise lenders on how to appraise space risks, and attorneys familiar with the compliance issues that space companies must navigate.

Let’s have a conversation.

Authored by Steven Kaufman, Alexis Sáinz, Randy Segal.