AI Week in Review - 8/9/25

Public Sector

Author

Chris Kraft
August 09, 2025

The Details

Federal

GSA and OpenAI have partnered to offer ChatGPT Enterprise to all federal agencies for $1 each, accelerating AI adoption and improving government efficiency under the America’s AI Action Plan.

GSA’s new OneGov agreement with AWS offers up to $1 billion in savings to modernize federal IT, accelerate cloud adoption, and advance AI innovation across government agencies.

GSA has added Anthropic’s Claude, Google’s Gemini, and OpenAI’s ChatGPT to its Multiple Award Schedule, expanding federal agencies’ access to advanced AI tools for transformative, mission-critical applications

Two Chinese nationals are charged with illegally exporting tens of millions in AI-capable microchips to China, allegedly using transshipment points to evade U.S. export laws.

NIST is developing a “cyber AI profile” built on its Cybersecurity Framework to address AI risks without overwhelming professionals, while lawmakers push for expanded AI testing and red-teaming standards.

The U.S. Navy is exploring AI-driven behavioral profiling to strengthen authentication in zero trust environments, binding users to devices and detecting anomalies for enhanced cybersecurity.

A CSIS report warns AI could lower barriers to bioterrorism, urging U.S. policymakers to strengthen safeguards for biological design tools, modernize synthesis screening, and fund NIST and CAISI to counter emerging threats.

State / Local

California is partnering with Google, Adobe, IBM, and Microsoft to deliver AI education and workforce training to over two million students, preparing them for high-paying careers in emerging tech.

Illinois has enacted a law banning AI from providing therapy or psychotherapy, ensuring only licensed professionals deliver mental health care while allowing AI for administrative support.

Bellevue pilots AI tools that streamline permitting, aiming to boost efficiency, support housing and economic growth, and enhance customer service with safeguards for ethical use.

California is partnering with Adobe, Google, IBM, and Microsoft to provide free AI tools, training, and career programs to students and teachers, preparing them for an AI-driven economy.

International

The U.S. and China’s AI action plans reveal competing global visions—U.S. market-driven openness versus China’s state-controlled governance—shaping international standards, tech access, and influence over Global Majority nations.

Chinese universities are embracing AI in education, integrating tools like DeepSeek into curricula to boost digital literacy, productivity, and job readiness—contrasting with Western caution over academic integrity.

Australia’s Digital Transformation Agency has released a government AI technical standard to embed transparency, accountability, and safety across the AI lifecycle, guiding ethical public sector adoption without stifling innovation.

Germany missed the EU deadline to appoint AI oversight authorities under the AI Act, prompting warnings from regulators and consumer groups about risks to innovation, consumer protection, and enforcement.

UK MP Mark Sewards has launched “AI Mark,” a digital twin built with Neural Voice to provide constituents 24/7 access for questions and local issues, aiming to boost political engagement.

South Korea is building a national open-source AI model using domestic chips, software, and talent, aiming for AI sovereignty and positioning itself as an alternative to U.S. and Chinese systems.

Everything Else

OpenAI has launched GPT-5, a major upgrade to its chatbot technology, offering faster, more accurate, and more humanlike reasoning capabilities, now powering the free version of ChatGPT.

Ex-Google exec Mo Gawdat warns AI will decimate jobs—including CEOs and government leaders—dismissing the idea that it will create work as “100% crap.”

OpenAI has released two open-weight reasoning models under Apache 2.0, offering near state-of-the-art performance, strong safety measures, and efficient deployment for developers, enterprises, and governments.

Stanford’s Erik Brynjolfsson warns AI could erase many entry-level jobs and widen inequality, urging investment in human-centered skills, fair wealth distribution, and managing economic transitions alongside rapid technological advances.

A new survey reveals a major AI adoption gap: while most executives see success, fewer than half of employees agree, citing poor training, low AI literacy, and workflow mismatches.