This Week in Government Tech Media – February 27, 2026

By Brad Bass /

Government cybersecurity and AI issues never fail to generate headlines, and this week was no different. We saw various stories on recent events at the Cybersecurity and Infrastructure Security Agency (CISA), including some major leadership changes. And we also saw a continuation of the disagreement between the Pentagon and AI company Anthropic. Here’s a roundup of those stories and more:

  • A CISA Shake-Up and Warning
  • The Pentagon/Anthropic Saga Continues

A CISA Shake-Up and Warning

CISA was featured in numerous headlines in the government tech media this week and not for the best of reasons. Several reporters covered personnel changes at the highest levels of the agency as well as a warning from the agency to federal organizations about a critical flaw in widely used networking software:

  • CISA’s controversial acting director Madhu Gottumukkala was expected to depart the agency for a new role amid news of several leadership shakeups and incidents during his tenure, with Executive Assistant Director for Cybersecurity Nick Andersen taking over as acting director, according to an article by David DiMolfetta in Nextgov/FCW.
  • In CyberScoop, Tim Starks quoted a Department of Homeland Security (DHS) statement on Gottumukkala’s departure that said he “tackled the woke, weaponized, and bloated bureaucracy that existed at CISA, wrangling contracts to save American taxpayer dollars.”
  • In a separate report, DiMolfetta noted that CISA CIO Bob Costello, whom Gottumukkala attempted to transfer out of the agency, has received multiple offers for his next job, though it’s not clear if those opportunities are with other government agencies or with industry.
  • The news on CISA’s leadership shakeup might not be surprising in light of an analysis by Starks in CyberScoop noting a large consensus, if not total unanimity, among those who have worked with and for the agency that it has suffered significantly since the arrival of the new White House administration.
  • In other CISA news, Justin Doubleday reported in Federal News Network that the agency issued an emergency directive this week instructing federal agencies to address critical vulnerabilities in some Cisco networking systems and evaluate whether any compromises have occurred.
  • Covering the directive for CyberScoop, Matt Kapko wrote that CISA said attackers have been exploiting a pair of zero-day vulnerabilities in Cisco’s software for at least three years, and the global campaign is ongoing.
  • In MeriTalk, Grace Dille quoted a statement from Gottumukkala noting that “the ease with which these vulnerabilities can be exploited demands immediate action from all federal agencies.” 
  • The directive came amid a multi-week DHS shutdown and put “additional strain on incident response and vulnerability management operations across the agency and its subdivisions,” according to coverage by Chris Riotta in GovInfoSecurity.

The Pentagon/Anthropic Saga Continues

As I noted last week, the Pentagon’s threat to designate Anthropic’s Claude AI platform as a supply chain risk has been covered extensively by the media in general and continues to grab the attention of government tech media outlets. We saw additional coverage of the story this week, as well as other stories illustrating how the Pentagon’s move to embrace AI continues unabated:

  • Axios reporters Dave Lawler, Colin Demarest and Maria Curi broke the news that Pentagon asked two major defense contractors to assess their reliance on Anthropic’s AI model, Claude – a first step toward a potential designation of Anthropic as a “supply chain risk.” 
  • Sources told Patrick Tucker of Defense One that it could take three months or even longer for the U.S. military to regain access to such a powerful tool on its classified networks if the Pentagon carries out its threat to blacklist the Claude AI platform.
  • Analysts said the Pentagon’s threat against Anthropic risks triggering unintended cybersecurity and supply chain consequences across the defense industrial base, according to an article by Riotta in GovInfoSecurity.
  • Lee Ferran and Sydney Freedberg reported in Breaking Defense that Anthropic’s statement that it will not agree to demands issued from the Pentagon regarding how its software can be used, despite the threats to punish the company.
  • Meanwhile, the Pentagon continues to move forward on multiple fronts in deploying AI, with Brandi Vincent of DefenseScoop writing that the military is looking to purchase AI-powered coding assets to facilitate multipart engineering tasks with little human intervention.
  • Lisbeth Perez reported in MeriTalk that the initiative aims to give DOD software developers commercial-grade AI capabilities to accelerate and improve software delivery.
  • Interviewed by Francis Rose for the “Fed Gov Today” program, the Pentagon’s Chief Digital and Artificial Intelligence Officer Cameron Stanley said GenAI.mil platform saw more than 550,000 unique users log on to use primarily frontier AI models in its first week alone and is now approaching a million users across the department.
  • In a related story, Jane Edwards noted in ExecutiveGov that the Navy officially designated GenAI.mil as its enterprise IT service for controlled unclassified information, or CUI, and Impact Level 5 generative artificial intelligence use.
  • The Air Force is investigating the use of AI for battle management, discerning the differences between “human-in-the-loop” and “human-on-the-loop,” where the latter is a more sophisticated role of the machine in decision-making, according to a piece by Kimberly Underwood for SIGNAL Media.

Upcoming Industry Events

As we enter March, we noted a few upcoming industry events you might want to check out:

If you would like your event included in this list, please fill out this form.

There’s a new episode of the podcast “Gov & Beyond” this week, and it’s a great one! This time, my colleagues Luca Pagni and Joyson Cherian interviewed Terry Gerton, the host of Federal News Network’s “Morning Drive” program. She’s had a fascinating career and shares her insights on the stories and issues that she’ll be covering this year.

That’s all for now. If someone forwarded this to you, you can cut out the middleman and subscribe to this newsletter on LinkedIn or via the form below.