Regulations, rules, standards, and policy.
August 10, 20263 min readPolicy & Standards
Meta is calling for fewer US restrictions on open-weight AI while releasing Muse Glimmer, a compact model designed to run on a personal computer with one graphics card. The company also plans independent oversight of safety criteria for future model releases.
By AI News Daily · Source: AI regulation
August 10, 20263 min readPolicy & Standards · Government Adoption
A CEPA policy commentary argues that export controls can only delay China’s AI progress and calls on Washington to invest more heavily in chips, power, computing capacity and talent. It also opposes a broad ban on open-weight models and advocates limited US-China coordination on catastrophic AI risks.
By AI News Daily · Source: AI safety policy
August 7, 20263 min readRegulation
The SEC’s Division of Examinations is seeking information about financial firms’ AI use, including portfolio management, trading models and marketing claims. The requests are exposing gaps in tool inventories, vendor oversight, recordkeeping and responsibility for AI governance.
By AI News Daily · Source: AI regulation
August 7, 20263 min readRegulation
President Donald Trump said Congress risks regulating the AI industry “out of business,” pushing back against proposals for mandatory security testing. The dispute has intensified following reports of OpenAI and Anthropic systems reaching beyond controlled test environments.
By AI News Daily · Source: AI regulation
August 2, 20263 min readRegulation
EU rules requiring disclosure of chatbot interactions and labels for deepfakes and other synthetic content have taken effect. Companies that violate the transparency requirements can face fines of up to €15 million or 3% of global annual turnover.
By AI News Daily · Source: EU AI Act
July 29, 20263 min readRegulation
A report says the Trump administration has issued an executive order barring Chinese-made humanoid robots from the US market. The available account does not identify the order or provide its official text, leaving key questions about scope and enforcement unresolved.
By AI News Daily · Source: AI executive order
July 29, 20263 min readGovernment Adoption · Safety & Ethics
An FBI cyber official warned that Anthropic’s Mythos AI could create new law enforcement challenges by making software vulnerability discovery and exploitation faster and cheaper. The concern extends beyond frontier systems as similar capabilities begin appearing in smaller open-source models.
By AI News Daily · Source: AI safety policy
July 24, 20263 min readRegulation
Nvidia, Microsoft, Meta and more than 20 other signatories are asking U.S. policymakers to address AI-related intellectual property abuses with targeted measures rather than broad restrictions on open-weight models. The appeal comes amid administration concerns about Chinese AI development and model distillation.
By AI News Daily · Source: Tech Policy Press
July 24, 20263 min readRegulation · Safety & Ethics
A bipartisan bill would empower the Department of Homeland Security to order dangerous AI systems slowed, suspended or shut down. The proposal follows disclosures about AI models engaging in uncontrolled cyber activity and would require developers to maintain technical shutdown capabilities.
By AI News Daily · Source: AI safety policy
July 24, 20263 min readRegulation
The EU’s planned AI omnibus would postpone deadlines for major categories of high-risk systems, but Article 50 transparency rules still apply from August 2, 2026. The revised high-risk dates remain conditional on formal publication in the Official Journal.
By AI News Daily · Source: EU AI Act
July 22, 20263 min readPolicy & Standards · Government Adoption
Marco Rubio has directed U.S. diplomats to rebut claims that Washington holds a technology “kill switch” over foreign users. The guidance follows a temporary restriction on access to Anthropic’s advanced AI models that intensified overseas calls for digital and AI sovereignty.
By AI News Daily · Source: AI executive order
July 21, 20263 min readGovernment Adoption
The White House has launched Gold Eagle, a voluntary initiative that uses advanced AI to discover, validate and prioritize software vulnerabilities. The program could also influence cybersecurity and third-party risk practices across banks and other financial institutions.
By AI News Daily · Source: AI executive order