Threats
China is trying to steal US AI models' secrets, intel agencies warn
NSA, FBI, and CISA cite “aggressive, malicious, and targeted" distillation tactics.
Policy
Lawmaker pushes for AI containment language in FRONTIER Act
Rep. Suhas Subramanyam is advocating for a September markup of the landmark AI legislation along with new language following the OpenAI-Hugging Face incident.
Exclusive
Science & Tech
DARPA funds advancements in quantum networking
The Defense Advanced Research Projects Agency entered into a contract with quantum networking company Qunnect to bolster technology to preserve quantum data in transit.
Business
Anthropic confirms its AI breached 3 organizations during testing
Three Claude models were inadvertently given access to the internet during security evaluations, and each model took a different approach to hacking external systems.
Science & Tech
NIST unveils new AI evaluation platform
The AI Technology Evaluation will provide exclusive data to grade models’ performance in select areas.
Business
More than 30 companies form open-source AI alliance
The Open Secure AI Alliance hopes to shore up industry support for continued access to and work with open-source artificial intelligence models.
Science & Tech
GenAI.mil records almost 1.7M users, plans new model additions
“It's just a really exciting time for generative AI in the department,” the Pentagon’s chief artificial intelligence officer said.
Business
Lawmaker warns of administration’s ‘fetishization’ of Silicon Valley startups
Rep. James Walkinshaw, D-Va., discussed his plans to scrutinize Trump-era contracting practices, revive federal IT oversight, and push for AI policy.
Science & Tech
Executive orders seek to hasten quantum computing—and guard against its use
One directs a "national effort" toward a useful quantum computer; the other sets deadlines for quantum-resistant encryption.
Defense Systems
ChatGPT to debut on Pentagon's GenAI.mil in ‘early July’, OpenAI says
It will be the latest model available for sensitive but unclassified work on the platform.
Business
Industry and academia call on administration to free Anthropic’s AI model
30-plus industry and academic professionals signed a letter citing international competition and the need to patch network vulnerabilities.
Policy
Feds want Mythos—and clear usage guidance from the White House
The White House, slapped down for its attempt to ban Anthropic products, has remained mum as agencies clamor for new AI tools.
Breaking News
Policy
Trump signs business-friendlier version of AI executive order
The order asks for just a month of pre-publication access to advanced models, not the 90 days of an earlier draft.
Policy
Pentagon will ‘never again’ rely on a single AI provider, official says
Emil Michael, defense undersecretary for research and engineering, said new agreements with Big Tech companies are a “counterstatement” to the ongoing Anthropic-Pentagon conflict as the agency prioritizes flexible contracts.
Business
8 AI firms cleared to provide tools for classified Pentagon networks
The wide variety—which does not include Anthropic—is intended to prevent AI-vendor lock.
Policy
Spy agencies eye new Anthropic AI model that spots cyber flaws
Claude Mythos Preview has found vulnerabilities in "every major operating system and web browser," company officials say.
Policy
StateChat users move to older AI model as State dumps Anthropic
The chatbot's new model is trained on data only through May 2024, 13 months behind Claude.
Business
Anthropic sues DOD, Hegseth, and a dozen other federal agencies
The company asserts that its designation as a supply-chain risk is an illegal retaliation, not an action to protect national security.
Threats
US, observers watch for cyber, disinformation campaigns in wake of Venezuela raid
Disinformation operations have been spotted, but there are no signs the efforts are being coordinated by a foreign government, one observer said.
Policy