What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Yes. OpenAI announced an agreement with the U.S. Department of Defense on February 27, 2026, to deploy its AI systems in classified government environments. The company published details the next day and said on March 2 that it was adding more explicit limits on domestic surveillance. On May 1, the Pentagon announced a wider classified-network AI rollout; OpenAI said its participation came under the agreement it had already announced, not a wholly separate deal.
The agreement provides for broad lawful defense uses, but OpenAI says it bars specified uses involving domestic surveillance, autonomous weapons and certain high-stakes decisions. Important details—including the full contract and how the restrictions are enforced—remain unavailable in the public descriptions.
What OpenAI agreed to
The agreement covers deploying OpenAI systems in classified Defense Department environments. It is not simply access to ordinary consumer ChatGPT, and the public information does not establish that the models are embedded in weapons or deployed across every military system.
Free tools Windows power users keep installed
One-click scans. No signup required.
OpenAI described the deployment as cloud-only, rather than running directly on edge devices. It said its safety stack would remain in place, with OpenAI-controlled classifiers used to monitor and update protections, and that cleared technical staff—including engineers and safety researchers—would support the work. Those are OpenAI’s descriptions of the arrangement; the public materials do not spell out every technical or operational detail. OpenAI’s announcement and subsequent update provide the company’s published account.
#1 Best Overall
What the agreement permits—and what it restricts
OpenAI said the Department could use the system for “all lawful purposes,” subject to applicable law, operational requirements and established safety and oversight protocols. That is broad authorization for lawful defense work, not permission to do anything the military wants. It also does not settle whether a particular use is wise, accountable or reliable simply because it is legal.
OpenAI identified three restrictions in its public description:
- Domestic surveillance: The system may not be intentionally used for domestic surveillance of U.S. persons or nationals. In its March 2 update, OpenAI said the language expressly covered deliberate tracking, surveillance or monitoring, including through commercially acquired personal or identifiable information.
- Autonomous weapons: The system may not independently direct autonomous weapons in circumstances where law, regulation or policy requires human control. This is not a blanket ban on every military use of AI or every autonomous capability.
- Other high-stakes decisions: The system may not make certain high-stakes decisions that require approval from a human decision-maker.
These limits leave room for AI-assisted military work. Potential categories include intelligence analysis, logistics, cybersecurity, mission planning and synthesizing battlefield information for people making decisions. The published restrictions do not amount to a general prohibition on targeting support, intelligence work or decision support. The key distinction is between assistance and a system taking an action or making a decision without required human authorization.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Why the March 2 surveillance change matters
OpenAI’s initial public framing referred to a prohibition on “mass domestic surveillance.” Critics questioned whether that qualifier left other forms of surveillance unclear. On March 2, the company said the agreement would be amended to state more explicitly that its systems could not be intentionally used to track, surveil or monitor U.S. persons or nationals, including by using commercially available personal or identifiable data.
OpenAI also said Department of War intelligence agencies such as the National Security Agency were excluded unless the parties reached a new agreement. That statement should not be expanded into a claim that every intelligence or national-security activity is excluded. The exact boundaries of incidental or non-targeted collection, legally authorized intelligence work and other edge cases are not fully resolved by the public summary. Senator Elizabeth Warren later sought the full contract language and asked specifically about incidental surveillance. Her letter illustrates why the public wording has continued to draw scrutiny.
How the deal followed the Anthropic dispute
The agreement came amid a public dispute between the Pentagon and Anthropic over restrictions on military use of Anthropic’s AI systems. Anthropic had objected to uses involving mass surveillance of Americans and fully autonomous weapons. OpenAI said its agreement preserved restrictions on those areas and added a safeguard for certain high-stakes automated decisions.
Rank #3
- Incredibly Light. Surprisingly Thin. - LG gram is designed to go wherever you do. Weighing just 2.5 lbs. with an ultra-slim 0.7-inch profile, it slips easily into your bag and feels light in hand—making it effortless to carry, commute, and work from anywhere.
- Remarkably Light. Reliably Strong. - LG gram has passed seven military-grade durability tests, striking an impressive balance between a highly portable, lightweight metal build and the confidence to handle everyday movement and travel.
- Power That Last with Smart Efficiency - LG gram combines a high-capacity 72Wh battery with AI-driven power management to optimize efficiency based on your usage. The result is up to 32 hours of video playback for} long-lasting performance that keeps up with your day—at home, at work, or wherever you go.
- AMD Ryzen AI Performance - Powered by AMD’s AI-optimized Ryzen processor with Radeon Graphics and a built-in NPU, LG gram delivers smooth multitasking and responsive performance. Fast 32GB LPDDR5x memory and 1TB NVMe storage keep everything moving without slowdowns.
- Dual AI for Always-On Intelligence - LG gram’s Dual AI—powered by EXAONE 3.5, LG’s AI solution—combines gram chat On-Device AI and gram chat Cloud AI to deliver seamless assistance. gram chat On-Device AI enables fast document search and summarization directly on your PC, while gram chat Cloud AI expands capabilities when connected—so everyday tasks stay smooth, responsive, and uninterrupted.
OpenAI also characterized its agreement as having stronger safeguards than previous classified-AI agreements, including Anthropic’s. But the full contracts have not been publicly compared in the material available here, so that claim should be attributed to OpenAI rather than treated as an independently verified, line-by-line comparison. The broader dispute is about who defines acceptable uses, whether contractual limits can meaningfully constrain government agencies, and how private vendors can oversee systems operating in classified settings.
Timeline: announcement, amendment and wider rollout
- February 27, 2026: OpenAI announced it had reached the agreement.
- February 28: OpenAI published its account of the arrangement and its safeguards.
- March 2: OpenAI announced additional, more explicit domestic-surveillance language and said intelligence agencies such as the NSA were excluded absent a new agreement.
- March 17: A separate AWS government-sales arrangement was reported, distinct from the core Pentagon agreement.
- May 1: The Pentagon announced agreements with technology companies to deploy AI on classified networks. OpenAI said its participation reflected the agreement announced in February.
The May announcement described deployments on classified Defense Department networks, including Impact Level 6 and Impact Level 7 environments, with stated aims such as data synthesis, situational awareness and augmenting warfighter decision-making. News coverage differed on whether the initial group contained seven or eight companies, with Oracle reportedly added after an initial list. That discrepancy concerns the broader Pentagon program, not the number of separate OpenAI-Pentagon agreements. The Associated Press reported that OpenAI described its participation as the earlier agreement.
What remains unclear
OpenAI has published selected language and its interpretation of the agreement, but the complete contract and implementation documents are not established in the public materials cited here. They do not provide a detailed public account of:
- Which models and applications are deployed, or the specific operational tasks they support.
- How audits, incident reporting, enforcement and remedies work if either party believes a restriction has been breached.
- How much visibility OpenAI retains into prompts and outputs inside classified systems, or how its classifiers function in practice.
- How model updates are approved and handled, or whether and how the terms address fine-tuned or derivative systems and outputs incorporated into other software.
- The precise treatment of incidental surveillance and other legally authorized collection.
OpenAI says its cloud architecture and safety stack enable it to verify compliance with its red lines. That is a claim about the intended safeguards, not public evidence that answers every question about their day-to-day operation or effectiveness. Classified environments also make independent public scrutiny difficult. Without the complete contract and operational details, outsiders cannot confidently assess how disputes would be resolved or what practical leverage OpenAI would have if it objected to a use.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the agreement is not
It is not a public authorization for unrestricted autonomous warfare. The published restriction is narrower: the system may not independently direct autonomous weapons where human control is required.
Recommended Free Tools
It is not a ban on all military AI or surveillance. The agreement allows broad lawful defense uses, while OpenAI says it prohibits intentional domestic surveillance of U.S. persons and nationals. The precise treatment of other surveillance scenarios remains a point of scrutiny.
Best Value
It is not a consumer ChatGPT arrangement. The deal concerns a specialized classified deployment; the public materials do not say ordinary ChatGPT conversations are transferred to Pentagon networks.
It is not the same as OpenAI’s AWS partnership. OpenAI separately announced an Amazon partnership, and reporting described an AWS arrangement to help sell OpenAI models to government customers for classified and unclassified work. That cloud distribution and government-sales relationship is distinct from the direct Pentagon agreement. See OpenAI’s Amazon partnership announcement and TechCrunch’s report on the AWS government-sales arrangement.
Why it matters
The agreement is a significant expansion of OpenAI’s defense work: it places the company’s systems in classified government environments while giving the Defense Department access for a wide range of lawful purposes. Its importance lies in the tension between that breadth and the vendor’s stated red lines. The public record identifies restrictions, but does not yet make their enforcement, technical reach or treatment of every borderline case easy to evaluate.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Scan for outdated or missing drivers - takes under a minute3Repair Windows errors before they cause bigger problemsOpenAI’s announcement uses the administration’s “Department of War” terminology; this article uses the conventional name, U.S. Department of Defense. The agreement was announced February 27, 2026, with details and later changes published by OpenAI.
Quick Recap
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.
