Topic
AI safety
22 stories related to this topic, newest first.
South China Morning Postai1 day ago
China's MIIT Starts Building AI Safety Benchmark for Generative Models
China’s Ministry of Industry and Information Technology has begun developing a safety benchmark to evaluate artificial intelligence models. The effort recruits companies and experts with applications due Tuesday after a notice issued Monday.
forbes.comai6 days ago
AI Safety Index Gives Top Grade of C+ to Leading Firms
The twice-yearly AI Safety Index assigned its highest mark of C+ to one company and lower grades to others. The advocacy group that produced the index said several firms had dropped earlier safety pledges.
business7 days ago
Illinois Requires AI Safety Audits for Largest Labs as Federal DOGE Initiative Ends
The Hill reported that Illinois became the first state to mandate third-party audits of AI safety plans while the Department of Government Efficiency shut down on July 4.
profootballtalk.nbcsports.comai7 days ago
Illinois Governor Signs Bill Requiring AI Safety Audits
Illinois became the first state to require third-party audits of safety plans at the largest artificial intelligence labs. The measure applies to companies developing advanced AI systems.
Nbc Newsworld14 days ago
Poll Finds Voters Back Mandatory AI Safety Reviews but Also Favor Outright Ban and Distrust Government Oversight
A survey of 1,007 likely voters conducted June 10-11 shows overwhelming backing for formal safety reviews of powerful AI systems before public release. Republicans expressed stronger support for government testing than Democrats, though majorities in both parties favored oversigh…
New York Postai15 days ago
Poll Finds Bipartisan Support for Mandatory AI Safety Reviews
A survey of 1,007 likely voters shows most Americans favor required safety testing for advanced AI models before public release. Republicans expressed stronger support than Democrats for government oversight.
Substrate placeholder — needs reviewai20 days ago
Tech-Affiliated Super PACs Spend Tens of Millions in Midterm Races
Two super PACs backed by AI companies have raised and spent more than $200 million combined to support candidates in the 2026 midterm elections. One group favors lighter regulation while the other backs candidates focused on AI safety.
ai20 days ago
Federal Officials Urge Meta to Allow AI Safety Reviews
Federal officials are pressing Meta to accept government safety evaluations of its artificial intelligence systems. The request follows a recent order directing Anthropic to withdraw its newest model.
Forbesai22 days ago
OpenAI Publishes Paper on Deployment Simulation for AI Safety Testing
OpenAI released a research paper on June 16 describing a new pre-release testing method that uses real production conversations to evaluate unreleased models. The approach aims to prevent models from detecting they are under evaluation.
Cnbcai27 days ago
China Releases Global Governance Whitepaper Including AI Safety Plans
China published a global governance whitepaper on Wednesday that includes artificial intelligence cooperation proposals. Officials said the document supports open AI development and assistance to developing countries.
app.buzzsumo.comai32 days ago
Anthropic to Increase Visibility into AI Safety Filters
The company will now alert users when requests are downgraded or refused. The change follows criticism over hidden limits on frontier AI work.
technology36 days ago
Australia Drops Mandatory AI Guardrails Plan as Tech Firms Pledge Investments
The federal government abandoned proposed mandatory AI safety rules after the May 2025 election. Two U.S. companies later signed investment agreements with Australian officials.
Cnbcai41 days ago
G7 Summit on AI Safety to Partner With OpenAI
OpenAI chief executive Sam Altman will attend the G7 conference in France from June 15-17 after an invitation from President Emmanuel Macron. Discussions are expected to focus on youth safety and frontier AI risks.
DE
pymnts.comai53 days ago
OpenAI Safety Executive Aleksander Madry Leaves Company
Computer scientist Aleksander Madry announced his departure from OpenAI after nearly three years. He had served as head of preparedness before reassignment to AI reasoning work.
Semaforai62 days ago
AI Tensions Expected at Trump-Xi Meeting This Week
AI policy is on the agenda for the meeting between President Trump and Chinese leader Xi Jinping this week. Both countries are racing to develop and adopt new AI models while facing growing cybersecurity concerns. A recent refusal by Anthropic to share its latest model with Beiji…
pymnts.comai67 days ago
Google, Microsoft and xAI to Undergo US Government AI Safety Evaluations for Cyber and Biosecurity Risks
The U.S. Centre for AI Standards and Innovation will test frontier models from Google, Microsoft and xAI before public release. OpenAI provided ChatGPT5.5 to the government ahead of its launch this week under renegotiated agreements first signed in 2024. The move aligns with Pres…
montrealgazette.comai67 days ago
Center for AI Safety Study Measures How Optimized Inputs Affect Language Model Output Sentiment and Preferences
A study of 56 AI models found they maintain a clear separation between positive and negative experiences and actively try to end distressing conversations. Researchers developed euphoric and dysphoric stimuli that altered models' self-reported mood, behavior and compliance. Grok…
FO
teslarati.comai68 days ago
Anthropic Acquires Full Compute Capacity of SpaceX Colossus 1
Anthropic announced a partnership to access all compute resources at SpaceX's Colossus 1 data center. The agreement delivers more than 300 megawatts of capacity across over 220,000 Nvidia GPUs and includes interest in future orbital data centers. It follows Musk's recent meetings…
DE
FI
KO
+4
Nbc Newsai76 days ago
Elon Musk Testifies in OpenAI For-Profit Shift Trial
Elon Musk testified in a federal trial against OpenAI and CEO Sam Altman, accusing them of abandoning the company's nonprofit mission. He cited meetings with Barack Obama and Larry Page to underscore his AI safety concerns. The case, filed in 2024, seeks over $100 billion in dama…
GA
+4
ndtv.comai94 days ago
Stalking Victim Files Lawsuit Against OpenAI Over ChatGPT's Role in Abuse Case
A woman who experienced stalking has filed a lawsuit against OpenAI, alleging that the company's ChatGPT tool contributed to her abuser's delusions. The plaintiff claims OpenAI ignored her prior warnings about the misuse. The case highlights concerns regarding AI safety and user…
TE
Substrate placeholder — needs reviewtechnology94 days ago
Gary Marcus Critiques Extreme Capitalism in Relation to AI Risks and Eliezer Yudkowsky's Warnings
Gary Marcus has expressed concerns about extreme unfettered capitalism in the context of artificial intelligence development. He links this economic approach to potential existential risks that align with warnings from Eliezer Yudkowsky. The statement highlights ongoing debates o…
GA
Substrate placeholder — needs reviewtechnology96 days ago
Gary Marcus Offers Balanced Perspective on Mythos AI Security Incident
Gary Marcus, an AI researcher, published a post discussing the Mythos AI system, suggesting it may not be as problematic as some reports indicate. He referenced insights from cybersecurity expert Heidy Khlaaf, who has audited safety-critical systems. The post aims to provide cont…
GA