The Next Orbital: Low Earth Satellites and AI are Changing the Fight against Corruption

Pairing low earth orbit (LEO) satellites with AI has led to groundbreaking gains in the fight against corruption, human rights violations, and environmental harm. But without safeguards and public oversight, this powerful new technology poses risks to civic space and security.

Credit: SpaceX via Unsplash

By Joseph Foti and Dieter Zinnbauer

Satellites carry numerous instruments that can be used for navigation, communication, and observation. Rapid changes in technology are making satellites cheaper and more accessible, particularly in the deployment of low earth orbit (LEO) satellites. When combined with artificial intelligence (AI), these systems may change how we approach international organized crime and repression.

This blog will focus on the potential benefits and risks of combining AI with LEO systems and provide recommendations for better safeguards and public oversight of this technology.

What’s Up? Big Changes in the Stars

In 2022, the US General Accountability Office estimated that there are 5,500 active satellites orbiting the earth. It predicted that that number would grow to more than ten times that size by 2030 — an additional 58,000 satellites.

While most people are familiar with Global Navigation Satellite System (GNSS) technology like GPS, attention is now moving toward low earth orbit (LEO) satellite arrays. These arrays are made up of hundreds or thousands of coordinated constellations (also known as “mega constellations”) — and the size of these arrays are growing. Some of the biggest LEO arrays include Starlink, OneWeb, Amazon Kuiper, China GW, Telesat Lightspeed, and Inmarsat Orchestra, all of which have plans to grow in size.

Table 1. Planned growth of LEO satellite arrays, by company

One of the main differences between GNSS satellites and the new LEO technologies is the altitude of each system. GNSS is in mid-Earth Orbit (between 2,000 and 22,236 kilometers above Earth), while LEO satellites are much closer to Earth (between 160 and 2,000 kilometers). This altitude affects which instruments can be used and how well particular instruments work, as detailed in the following bullets.

• Latency: Signals take time to travel between space and Earth. LEO satellites provide faster communication due to lower altitudes, while GNSS systems are optimized for precise positioning over longer distances.
• Primary Use: GNSS focuses on navigation and timing, while LEO constellations excel at communication and observation.
• Coverage: GNSS systems offer continuous global coverage with fewer satellites, whereas LEO systems require many more satellites to cover the globe.
• Signal strength and energy use: Because LEO satellites are closer to Earth, they need less energy to transmit information. This reduces costs per unit, although more units are needed since they cover less territory.

For the foreseeable future, large countries (and the EU) or major private investors will be the ones that can invest in and exploit these new technologies. Costs are likely to continue to fall, however, and the ensuing data and tools may become more widely shared. As a consequence, new countries can begin to use space and the number of applications will certainly multiply.

Picture This: How Optical Instruments and AI will Enhance LEO Systems

Advances in AI will also impact how LEO satellites are developed and used.

Optics and sensing technology (like infrared and radar systems) are improving significantly already, and AI is set to accelerate this trend. Resolution is increasing as well, according to the OECD. With the addition of AI, data scientists can synthesize this data to gather information on moisture, temperature, and topography. This should be by-and-large helpful for understanding human settlement, climate, and land use.

But what happens when the same technology is used for surveillance from space? Facial recognition and hi-resolution images of people, vehicles, or species may not be available yet, but its availability is a matter of “when,” rather than “if” as AI development continues.

The Good

Satellites and AI can enhance the rule of law, democracy, and science. They offer transformative possibilities for anti-corruption efforts, particularly in enforcing sanctions, monitoring contraband, and tracking human rights violations.

Sanctions enforcement

• One of the most promising uses of LEO satellite technology is in sanctions enforcement. From Russia to Venezuela, corrupt kleptocracies are fueling geopolitical tensions. One response has been to enforce international sanctions. Imagery and geodata shows that British ships are being used to evade sanctions on Russian oil. One use of satellite imagery is to enhance already existing geopositioning data to track sanctions evasion.
• AI-enhanced satellites can also monitor the movement of sanctioned goods — such as weapons, oil, or illicit finance — with greater accuracy and speed. By using real-time data to flag suspicious activity, governments and international bodies can act faster to prevent sanction breaches, enhancing accountability for nations and corporations alike.

Contraband

Illegal trade in natural resources — such as timber, minerals, and wildlife — is a multibillion-dollar black market that fuels corruption, conflict, and environmental destruction. AI-powered LEO satellites can better pinpoint deforestation hotspots, illegal mining, and smuggling routes. These systems can distinguish between legal and illegal logging operations or identify unregistered mining sites, providing actionable intelligence to governments and NGOs working to protect the environment and local communities.

Monitoring human trafficking, forced labor, and land grabs

The same technologies can also be used to tackle human rights abuses. Satellite imagery can reveal evidence of illegal land grabs and forced relocations and identify facilities suspected of using slave labor. By monitoring changes in settlement patterns, deforestation or wildcat mining near indigenous lands, and unusual construction activity in remote areas, AI tools can help authorities, companies along the supply chain, and civil society organizations to pinpoint sites of exploitation. These insights, in turn, can lead to more effective interventions.

Enabling response to environmental crimes

Tracking deforestation was one of the earliest applications of remote-sensing tools. With the help of AI, these systems turn from ex-post accountability to predictive warning systems. For example, before AI, researchers could only show that Brazilian authorities reacted only to one percent of illicit deforestation incidences. Now, researchers can track road building to better anticipate imminent deforestation and cut alert times by up to 50 days.

Emissions accountability

Tackling climate change is another application hotspot. Remote monitoring of source-level methane or CO2 emissions is foundational for a trustworthy climate governance system, where nationally determined contributions and reporting require robust measurement regimes. Yet such mechanisms have only become feasible at scale with the pairing of AI and advanced satellite technologies. This combined technology has already compellingly demonstrated its potential to detect corporation-level underreporting of emissions, in a case from the United States. Relatedly, as carbon sequestration becomes a growing business, ensuring that offset markets do not engage in fraud or regulatory manipulation will benefit from satellite monitoring.

These are a few of potential applications. Safeguarding fisheries, preventing piracy, and protecting human migrants are a few other applications, closely associated with corruption and transnational organized crime. What many of these applications have in common is that they provide novel sources for transparency, often at a highly detailed local level, beyond the immediate control of local or national authorities that might be implicated in these offenses and thus have long been able to suppress other sources of transparency and accountability within their area of influence.

The Bad

While the potential of these technologies is immense, they also come with serious risks, particularly in the hands of governments or private actors who can use them for repression or illicit purposes.

Surveillance

Satellites have long been used for surveillance, but LEO satellites combined with AI open new doors for governments to monitor citizens within their borders and those living abroad. This poses a particular danger in authoritarian regimes where dissent is criminalized, impunity for extrajudicial killings abounds, and there are few checks on search and seizure. For example, AI might be used to analyze movement patterns to identify protest gatherings. High-resolution imagery, combined with facial recognition technology, may at some point be able to pinpoint and track the location of political dissidents. The richest repressive regimes may also export this technology to allies to extend their surveillance networks beyond their borders.

Exporting terror and silencing dissidents abroad

Beyond the surveillance of local populations, there is growing concern about the ability of authoritarian regimes to track and silence dissidents who have fled to other countries. With global monitoring capabilities, even those in exile may not escape the reach of state repression. In a world where data can be traded between countries, LEO satellites could provide authoritarian governments with a detailed map of political opposition networks abroad, enabling harassment, intimidation, or worse.

Security threats

As with all advanced technologies, the militarization of LEO satellites and AI raises significant security concerns. These systems can be hacked, hijacked, or weaponized, putting critical infrastructure and communications in peril. The proliferation of satellite technology could also spark an arms race in space, as nations compete for control over increasingly valuable orbital assets.

Boosting the Good, De-Orbiting the Bad: Safeguards and Public Oversight

Strong safeguards and public oversight mechanisms are crucial to ensure that these powerful technologies are used responsibly. Given the newness of the subject, there is a need for model policies or rules, as well as convening around the topic. There are numerous possibilities for further work.

Clear public purpose and uses

Governments should establish clear guidelines that define the public purpose of satellite and AI systems, ensuring that their deployment aligns with human rights and anti-corruption objectives. Transparency about these purposes must be maintained through regular reporting and independent audits to prevent misuse for surveillance or repression.

Multi-sector oversight

In addition to government oversight, civil society and independent bodies should have a role in monitoring the use of these technologies. Multi-sector oversight committees can evaluate whether LEO satellite systems are being used ethically and whether their impacts align with public interest goals.

Human rights in public-private partnerships

As much of the LEO satellite infrastructure is privately owned, any partnership between governments and corporations must include strong human rights clauses. Governments and international organizations should ensure that private operators are not complicit in violating civil liberties or facilitating repression through the sale or misuse of satellite data. Where consistent with international law, such obligations should also limit weaponization.

Open data

Open geospatial data and satellite imagery have been central to open government approaches, from election results and infrastructure project monitoring to public health and reforestation initiatives. Making sure that the public can build applications off of such data (with appropriate privacy safeguards) will be essential to future growth. Public funds for satellites should go toward public science and data. This includes data that has been essential for open source intelligence and other movements that have worked against autocrats worldwide.

Open contracting and operations

There should be public transparency about who is operating in national and international airspace. Where governments license airspace and space on the radio frequency spectrum, such contracting should be open, competitive, and subject to scrutiny. Licensing and regulation of LEO satellite constellations should be made publicly available in open data formats, allowing watchdog organizations to monitor where, how, and for what purpose satellites are being used. By democratizing this information, citizens and civil society organizations can hold governments and corporations accountable.

Privacy protections and data security

States must have the capacity to regulate their aerospace and their radio frequency spectrum or cooperate for joint regulation. Where such agencies exist, we recommend the following.

• International cooperation: International cooperation should remain public and values-based. Democratic governments should not engage or contract technology companies or governments with values that are incompatible with democratic governments.
• Encryption: Sensitive personal or security information should have appropriate anonymization and encryption protocols, consistent with international human rights.
• Complaints process: Data protection authorities and cooperating agencies need the mandate to receive, investigate, and act on complaints. They should also be able to independently initiate and publicize their own investigations into incidents and patterns of practice.

Conclusion

The integration of LEO satellites and AI has led to groundbreaking tools in the fight against corruption and human rights violations. But without strong oversight, there is a real risk that these technologies could be used to undermine the very freedoms they seek to protect. Responsible governance, transparency, and an unwavering commitment to human rights are essential to realizing the full potential of this new frontier in anti-corruption.