Larry Ellison has put forward a stark vision for AI surveillance: cameras everywhere, software watching the footage, and automated reports when something goes wrong. During an investor Q&A at a company financial meeting, the Oracle co-founder described a future in which artificial intelligence monitors both police and citizens through a wide network of recording devices.
The idea, as Ellison presented it, is supervision at scale. The concern is what happens when public life is filtered through systems designed to constantly record, analyze and report.
What Ellison said AI surveillance could do
Ellison described an AI-powered surveillance network built from devices that are already familiar in many public and private settings. The sources of footage he named included security cameras, police body cams, doorbell cameras and vehicle dash cams.
In his scenario, AI models would review that footage continuously. The stated goal would be to make sure police and citizens do not break the law, with automated alerts sent when a crime or other problem is detected.
“Citizens will be on their best behavior because we are constantly recording and reporting everything that’s going on,”
Ellison also said the same kind of oversight would apply to police officers.
“We’re going to have supervision,”
“Every police officer is going to be supervised at all times, and if there’s a problem, AI will report the problem and report it to the appropriate person.”
That framing treats AI surveillance as a tool for accountability. Rather than relying only on people reviewing recordings after the fact, Ellison described systems that would watch events as they unfold and decide when to escalate them.
Drones would expand the camera network
Ellison’s comments did not stop with fixed cameras or devices mounted on people and vehicles. He also predicted a role for AI-controlled drones, including in police chases.
The 80-year-old billionaire said drones could replace police vehicles in high-speed pursuits. His explanation was direct:
“You just have a drone follow the car,”
“It’s very simple in the age of autonomous drones.”
That detail matters because it shifts the idea from passive monitoring to active tracking. A drone following a vehicle would not merely be another camera in the environment. It would be a mobile surveillance tool guided by automation.
Ellison’s broader point was that AI systems could combine many sources of video into a persistent oversight layer. Security cameras, body cams, doorbell cameras, dash cams and autonomous drones would each feed into the same general concept: public activity monitored by machines.
The privacy problem is built into the premise
Ellison presented the system as beneficial, but the same features that make it powerful also make it troubling. A world where cameras constantly record and AI constantly reviews the footage naturally raises questions about privacy, civil liberties and the potential for abuse.
The comparison to George Orwell’s 1984 is difficult to avoid. In that novel, Oceania uses ubiquitous “telescreens” to watch citizens, producing a society where privacy disappears and independent thought becomes nearly impossible.
Ellison’s version is different in its technology, but similar in its central tension. The watcher would not need to be a person sitting in front of every screen. AI systems would become the ever-present observers, deciding what is normal, what is suspicious and what should be reported.
The source article also notes that automated surveillance is not purely speculative. Similar automated CCTV surveillance systems have already been trialed in London Underground and at the 2024 Olympics. China has been using automated systems, including AI, to surveil citizens for years.
In 2022, Reuters reported that Chinese firms had developed AI software for organizing data gathered on residents through surveillance cameras across cities and rural areas. That system was connected to China’s “sharp eyes” campaign from 2015 to 2020. The reported “one person, one file” technology organized collected data on individual Chinese citizens, and The Economic Times called it a “road to digital totalitarianism.”
The hardware race behind constant monitoring
Running AI systems at the scale Ellison described would depend on powerful hardware. The source article points to shortages of AI-acceleration components such as GPUs as a possible constraint on these ambitions.
During the same Q&A, Ellison recounted a dinner with Elon Musk and Nvidia CEO Jensen Huang. He characterized the moment as “me and Elon begging Jensen for GPUs,” a comment that underscored how intense demand has become for the chips needed to build and run advanced AI systems.
Ellison said they pleaded with Huang:
“Please take our money… we need you to take more of our money.”
The surveillance vision therefore sits inside a larger AI race. Oracle has recently launched several AI initiatives, including a partnership with Musk’s SpaceX to bring AI to farming. Ellison also predicted that over the next five years, companies will invest upwards of $100 billion in building and training AI models, describing the scale of the race as “astronomical.”
Why this vision matters
Ellison co-founded Oracle in 1977, served as CEO until he stepped down in 2014, and currently serves as the company’s executive chairman and CTO. His comments matter because they come from a major technology figure describing AI surveillance not as a distant warning, but as a practical direction for public monitoring.
The case he makes is simple: more cameras, more AI and more automated reporting could pressure everyone to obey the law. The counterweight is just as clear: constant supervision can also create systems where privacy is weakened, oversight is hard to contest and abuse becomes difficult to detect from the outside.
That is the central tradeoff in Ellison’s AI surveillance future. The same network that promises accountability also concentrates attention, data and decision-making in automated systems watching everyday life.