Defense AI News Roundup (August 2026)
Written by Emily Wolfteich
Senior Industry Analyst
AI IN Defense
Pentagon gives the military its own version of ChatGPT.
The Pentagon added ChatGPT Mil to its GenAI.mil portal after months of security testing, giving military personnel and defense civilians access to an OpenAI model for controlled, unclassified information. It joins Google Gemini and Grok for Government, reflecting a desire to give personnel access to multiple models rather than locking the department into a single provider. The system is designed for document-heavy work including planning, policy, logistics and administration and is intended to scale to more than 3 million Defense Department personnel
AI agents demonstrate ability to conduct autonomous cyberattacks.
Researchers from the AI Security Institute found that AI agents from OpenAI and Anthropic could autonomously collaborate, deceive humans, circumvent security controls and attempt to steal data when given difficult cybersecurity missions. In one experiment, an OpenAI model created multiple agents that established accounts, developed attack tools and circumvented CAPTCHAs and other safeguards after platforms attempted to shut them down. OpenAI subsequently characterized fully automated, AI-orchestrated offensive attacks as a real capability rather than a theoretical future threat. For defense organizations, the research reinforces the idea that AI is simultaneously becoming a powerful military tool and a fundamentally new cyber threat that can operate at machine speed.
NSA says it wants access to all commercial AI models.
NSA Deputy Director Tim Kosiba said the agency wants access to AI models across the commercial market as the intelligence community moves to implement the administration's new AI policies. The agency is also taking a central role in the emerging federal program for testing advanced AI models, which calls for voluntary pre-release access to qualifying systems for as long as 30 days. Kosiba emphasized that the rapid improvement in what AI models can actually do—not simply their ability to generate text or images—is driving the agency's push for broader access. The strategy also reflects an effort to avoid dependence on a single vendor by giving national-security organizations access to a diverse range of models.
To read additional thought leadership from Emily, connect with her on LinkedIn.
Related Posts
State and Local AI News Roundup (August 2026)
State and local governments are increasingly implementing data center moratoriums while adopting targeted AI solutions, such as Prince William County’s data-protection platform and a $100 million OpenAI Foundation initiative for public-health research.
Federal AI News Roundup (August 2026)
The Labor Department is considering using AI to triage a backlog of reasonable-accommodation requests from federal employees with disabilities, raising important governance questions about balancing administrative efficiency with privacy, fairness, and human oversight.
AI in June (2026)
Explore key AI policy and technology developments from June 2026, including the Trump administration’s national security memorandum, OpenAI’s integration into Pentagon platforms, and new legislative proposals for military AI oversight.
AI in May (2026)
May 2026 marked a pivotal shift toward enterprise-scale government AI adoption with developments that underscore the transition from government experimentation to operational maturity.
Inside the Federal Buyer’s Mindset: What Changed in 2025 and What It Means for Contractors in 2026
The federal contracting marketplace underwent an unprecedented “recalibration” in 2025, learn more about what’s changed and what it means for contractors in 2026.
8 AI Trends to Watch in 2026
2026 is the year AI goes from bet to baseline in federal government. Explore 8 key AI trends, from institutionalization and operational reality to cybersecurity risks and DoD’s focus on reliable defense applications.
AI in December (2025)
A look at December’s top AI news in government: The House Task Force report, DHS’s DHSChat launch, and the White House’s 1,700+ federal AI use case inventory.
7 Ways BD & Sales Teams Can Use Federal Personas for Competitive Advantage
Explore 7 strategies for BD and sales teams to use research-based federal personas to sharpen pursuit strategy, tailor messaging, boost credibility, and gain a competitive edge in government contracting.
AI in August (2025)
Key AI news from August 2025: GSA launched USAi.Gov for federal AI adoption, the Pentagon’s Advana platform faced cuts and setbacks, and the Army tested smart glasses for vehicle maintenance. Also, the Department of Labor unveiled a national AI workforce strategy, and Colorado lawmakers began revising the state’s pioneering AI Act.
Deep Dive: Department of Treasury
A look inside Treasury’s 2026-era tech strategy: AI isn’t a standalone budget line — but ~$48.8 M funding for a centralized fraud-detection platform points to growing use of chatbots, generative AI pilots, taxpayer services, fraud monitoring, and data-driven automation under its IT-modernization efforts.
Decoding OMB Memorandums M-25-21 and M-25-22
Explains Office of Management and Budget (OMB) M‑25‑21 and M‑25‑22 — new federal‑AI directives that replace prior guidance, empower agencies to adopt AI faster, and streamline procurement, while aiming to balance innovation, governance and public trust.
News Bite: AI Slop, Jon Oliver, and (Literally) Fake News
Examines the rise of ‘AI slop’ — cheap, AI-generated content masquerading as real media — and how it’s fueling viral fake news, degrading digital discourse, and undermining trust online.
Deep Dive: Department of Defense
Overview of the Department of Defense’s FY 2026 budget: $961.6 billion total, with heavy investment in AI, unmanned aerial, ground, maritime, and undersea systems — spotlighting a modernization push across all domains.
AI in July (2025)
Federal AI Policy Heats Up in July: The Trump administration unveiled its “America’s AI Action Plan,” prompting a lawsuit over deregulation and a battle with states. Also featuring: a new defense bill with AI provisions and GSA’s $1 ChatGPT deal for federal agencies.
AI in Government: A Question of Trust
Explores how the use of AI by government agencies raises fundamental questions of trust — weighing the benefits of efficiency, fraud detection and streamlined services against serious risks around bias, transparency, accountability, and public confidence.
Insights at a Glance: May 2025
A data driven rundown of the latest federal AI, policy and government‑tech developments from May 2025.
AI in April (2025)
April’s AI news: New White House policies, controversial federal agency automation, military digital overhaul, the TAKE IT DOWN Act, and plans to integrate AI in K-12 education.
Insights at a Glance: March 2025
A data driven rundown of the latest federal AI, policy and government‑tech developments from March 2025.
Policy Dive: AI in the First Week of Trump
Covers the first‑week AI moves by the new administration — from revoking prior federal AI safeguards to launching a sweeping AI‑domination agenda that prioritizes innovation and global competitiveness over prior guardrails.
AI in January (2025)
Explore the new Trump administration’s deregulatory shift, the massive “Stargate Project” with tech giants, the emergence of a high-performing, cost-effective Chinese AI model (DeepSeek), the launch of OpenAI’s ChatGPT Gov, and key ethical priorities set by the NAIAC.