Thales chair and CEO Patrice Caine says AI risks should be addressed through technical safeguards, government oversight and an international framework—not by relying on technical controls alone. Reuters reported his call on September 17, 2026. The phrase “AI pause is no solution” is headline framing; Reuters did not report it as Caine’s verbatim wording.
What did Caine say about pausing AI?
In remarks to reporters at a presentation of war-fighting technology in Paris, Caine argued that AI can be controlled with technical measures, but that governance belongs to governments. Reuters quoted him saying: “I am convinced that we have the ways and means to have AI under control. It’s not a technical debate, it falls in the hands of governments.” He added: “The world needs a proper framework that comes on top of these technical guarantees.” Reuters, September 17, 2026
As an Amazon Associate I earn from qualifying purchases.
Those remarks support a policy position centered on safeguards and oversight. They do not establish that Caine used the exact words “AI pause is no solution”; the phrase is a summary in the headline framing, not a verified direct quote.
What safeguards and rules does he support?
Caine’s proposal combines two layers: technical guarantees intended to manage risks in AI systems, and public authority to set rules for development. He called for an international framework in addition to technical measures. The Reuters report does not lay out a detailed treaty, enforcement mechanism or timetable, so the proposal is best understood as a call for coordinated governance rather than a specific policy blueprint.
#1 Best Overall
The distinction matters: safeguards inside a system may address particular technical hazards, while governments determine the rules and accountability around development and use. An international framework would extend coordination beyond a single company or country. Caine’s remarks do not prove that any one layer is sufficient or that existing safeguards are effective in every setting.
How does human control work in Thales’s defense examples?
In a February 12, 2025 interview with McKinsey, Caine described AI applications in defense sensors and decision-support systems. AI can help distinguish relevant detections in radar data. In command-and-control systems, software can evaluate complex mission factors and propose a course of action for an operator to assess. Caine said: “AI algorithms have demonstrated that they add real value by quickly finding the best possible compromise—and then leaving it up to the operators to accept or reject what the AI proposes.” McKinsey, February 12, 2025
In this account, human control means an operator retains the decision to accept or reject an AI-generated proposal. It is not the same as eliminating the system’s influence: the AI still filters information and shapes the options presented to the operator.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What risks remain even with an operator in the loop?
Caine has identified several practical constraints in describing Thales’s work. Operational data may be classified and therefore not exportable; training data needs to reflect the context where a system will be used; and cyberattacks can affect how an AI system interprets information. These are his descriptions of engineering and security concerns, not independent evidence that a particular safeguard prevents them.
- Data restrictions: Classified operational information can limit which data may be used or transferred.
- Context-specific training: A system trained on data that does not reflect its deployment setting may not interpret conditions as intended.
- Cybersecurity: An attack can potentially manipulate or disrupt interpretation, even when a human operator makes the final decision.
These concerns show why human approval and technical safeguards are complementary rather than interchangeable. An operator can reject a recommendation, but the quality of the information and recommendation still matters.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Why does Thales’s perspective matter?
Caine is speaking as the head of a defense and technology company, where AI has both operational applications and governance implications. Reuters described Thales’s Hexaforce as an AI-based command-and-control system and reported the company’s characterization of it as sovereign and free from US ITAR export controls. Those are descriptions attributed to Thales, not independent findings about system performance or export-control status. Reuters, September 17, 2026
A separate Reuters-sourced report by CNA on October 1, 2026 covered Caine’s comments that AI-enabled cyberattacks are accelerating and that defenders can use AI at comparable scale and speed. It also reported Thales’s presentation of cybersecurity tools, including software protection against AI-powered attacks and a Google Cloud partnership. That cybersecurity discussion is related context, but it does not establish that his earlier call for an international framework was a product pitch. CNA, October 1, 2026
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
