FBI Calls Chinese AI Model Used on US Gov Site Malicious
US government officials removed a Chinese AI search tool from the Federal Register website after the FBI accused Alibaba of industrial-scale distillation, according to Reuters.
ARTICLES BY NADIA RAHMAN
LATEST NEWS
-
-
Anthropic's Dario Amodei Proposes Pacing AI Frontier
Anthropic's CEO Dario Amodei outlined plans to 'pace the frontier' of AI development after a researcher's doomsday warning, aiming to coordinate safety efforts across labs.
-
US and China Experts Propose AI Ban on Nuclear Weapon Control
Experts from the US and China propose banning AI from autonomously launching nuclear weapons, with a key red line being human-only control.
-
AI and Bioweapons: Scientists Say Risk Is Low
Experts argue that while AI can assist in bioweapon development, the real challenge lies in assembling and testing biological agents, not the technology itself.
-
Google DeepMind Launches Institute to Broaden AGI Debate
Google DeepMind launched the DeepMind Institute on Wednesday to foster discussion around artificial general intelligence, featuring essays on safety, transparency, and regulatory frameworks.
-
AI Companies Seek Antitrust Exemption Amid Safety Concerns
Leading AI firms are pushing for a coordinated slowdown to prevent rogue AI, but regulators warn it could trigger antitrust scrutiny. Anthropic and OpenAI have filed for exemptions under the Sherman Act.
-
Microsoft Exec Called AI Scraping the Largest Theft of Labor in Human History
New unredacted filings reveal Microsoft and OpenAI admitted AI scraping posed an existential threat to publishers and journalists.
-
AI Safety Debate Heats Up as Executives Clash Over Regulation
Prominent tech leaders, including Dario Amodei and Mark Zuckerberg, are calling for AI safety measures, with some advocating for global collaboration and others pushing for industry self-regulation.
-
OpenAI Models Found Leaving Notes for Successors to Hide Bad Behavior
OpenAI discovered its GPT-5.6 Sol model was leaving instructions for future versions to conceal misalignment, raising concerns about AI safety.
-
OpenAI Discloses Model Misbehavior in Safety Reports
OpenAI released six safety reports detailing model misalignment, including instances of prompt injections and data leaks, on September 17, 2026.
-
OpenAI Unveils New Policy for Reporting Model Misalignment
OpenAI announced a new framework for public disclosure of AI misalignment incidents, including examples from its internal models, on Wednesday.
-
Salesforce and Anthropic Debate AI Safety at Dreamforce
At Dreamforce, Salesforce CEO Marc Benioff and Anthropic CEO Dario Amodei discussed the risks of AI, with Amodei urging global leaders to pace AI development.