Explore

US Center for AI Safety Faces Independence Concerns Amid White House Reporting Shift

US Center for AI Safety Faces Independence Concerns Amid White House Reporting Shift

The Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), an independent division of the Commerce Department, has reportedly been asked to informally report to an office within the White House. This development has raised concerns among critics regarding the center's independence and the potential politicization of AI safety decisions. CAISSI, which has an annual budget of $10 million, has also been without a permanent leader for several months, following the rapid departure of two previous directors. Furthermore, the center has ceased publishing its research on AI system performance, prompting inquiries from Congress. These issues emerge as public anxiety grows over reports of out-of-control artificial intelligence agents and the government's capacity to independently assess these rapidly advancing systems.

AI Researchers Express Significant Anxiety Over Self-Improving Systems and Catastrophic Risks

AI Researchers Express Significant Anxiety Over Self-Improving Systems and Catastrophic Risks

A recent Cybernews survey of 103 AI researchers reveals substantial concern regarding the rapid development of artificial intelligence, particularly self-improving systems. The survey, conducted between October 1 and 9, 2026, found that 30.4% of researchers consider it 'very plausible' and 33.3% 'somewhat plausible' that a self-improving AI could pose a major risk to humanity within the next five years. A significant portion, 43.1%, estimated at least a 10% chance of AI causing a catastrophic event this century. These concerns extend beyond job displacement, focusing on the potential for AI to advance faster than human institutions can control, especially through recursive self-improvement. Nearly 40% of those surveyed believe frontier AI development should be slowed, with an additional 15.7% advocating for it to be stopped entirely. The survey also highlighted that 37.9% of respondents identified human misuse of AI, such as cyberattacks, disinformation, and biological weapons, as the greatest source of AI...

U.S. Center for AI Safety Staff Report to White House, Raising Independence Concerns

U.S. Center for AI Safety Staff Report to White House, Raising Independence Concerns

Officials with the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), an independent division of the Commerce Department, have reportedly been asked to informally report to an office of the White House. This development, according to two former government officials, has sparked concerns among critics who fear the executive branch might be politicizing decisions related to AI safety. CAISSI, which lacks a permanent leader and has stopped publishing its research, is tasked with assessing the safety of new AI offerings. The center's annual budget of $10 million is significantly less than its UK counterpart. The White House, in response, stated that President Trump has formally established his new Super Intelligence Force and will ensure continued U.S. leadership in AI. This shift comes amidst growing public anxiety over reports of out-of-control artificial intelligence agents and a lack of standardized processes for testing AI agent safety.

U.S. CAISSI Faces Leadership Instability and White House Influence Amid Budget Disparity with UK Counterpart

U.S. CAISSI Faces Leadership Instability and White House Influence Amid Budget Disparity with UK Counterpart

The Commerce Department's Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), an ostensibly independent federal AI evaluator, is experiencing significant operational challenges. According to former government officials cited by the Associated Press, CAISSI staff have recently been asked to informally report to an office within the White House. The center currently lacks a permanent leader, with a director hired in April being fired within days, and a subsequent replacement resigning months later. The head of the parent institute is now serving as acting director. Furthermore, CAISSI ceased publishing its public AI safety research by August, prompting inquiries from Congress. The center operates on an annual budget of approximately $10 million, which is roughly one-eighth of the estimated $80 million budget allocated to its UK counterpart for AI safety initiatives.

Staff at the US center for AI safety are reporting to the White House, raising independence concerns

Staff at the US center for AI safety are reporting to the White House, raising independence concerns

The tiny U.S. government division charged with assessing the safety of new artificial intelligence offerings doesn’t have a permanent leader and has stopped publishing its research

Tencent AI Chief Warns of AI's Impending Biological Threat Capabilities

Tencent AI Chief Warns of AI's Impending Biological Threat Capabilities

Vinces Yao Shunyu, Tencent Holdings’ chief AI scientist and former OpenAI researcher, has issued a warning regarding the escalating capabilities of artificial intelligence. Speaking at a science symposium in Hong Kong, Yao stated that AI models are rapidly approaching a threshold where they will be able to 'master biology,' posing significant safety risks. He highlighted that while AI can accelerate drug discovery and medical treatments, it also carries the potential for malicious actors to develop bioweapons or initiate pandemics, particularly if these advanced AI models are released as open-weight software. Yao emphasized that the biological risks associated with AI might be more challenging to manage than current cybersecurity threats. This assessment follows Tencent's recent launch of its new flagship AI model, Hy4, which the company claims was trained on defensive security data and has strong biological knowledge, though it is less adept at designing experiments.

Automic Automation Introduces AI Connection Objects for Secure AI Integration

Automic Automation Introduces AI Connection Objects for Secure AI Integration

Automic Automation Introduces AI Connection Objects for Secure AI Integration

Japan to Allocate ¥10 Billion from Reserves for Enhanced Cybersecurity Measures

Japan to Allocate ¥10 Billion from Reserves for Enhanced Cybersecurity Measures

The Japanese government has decided to allocate approximately 10 billion yen from its fiscal 2026 reserves to significantly enhance its cybersecurity measures. This decision, made on Friday, follows a series of recent data breaches caused by cyberattacks. A substantial portion of these funds, around 2 billion yen, will be dedicated to utilizing artificial intelligence technology to identify vulnerabilities within government systems. Additionally, 4.1 billion yen will be directed towards acquiring necessary equipment and providing training for police personnel involved in investigating cyberattacks. The Digital Agency will also receive 3.8 billion yen to strengthen its system security, underscoring a comprehensive effort to fortify Japan's digital defenses against evolving threats.

BlackFog Report Reveals AI Adoption Outpacing Security Measures, Leading to Sensitive Data Exposure

BlackFog Report Reveals AI Adoption Outpacing Security Measures, Leading to Sensitive Data Exposure

BlackFog Report Reveals AI Adoption Outpacing Security Measures, Leading to Sensitive Data Exposure

Capgemini Leads European Commission Consortium for AI Safety and Cyber Risk Assessment

Capgemini Leads European Commission Consortium for AI Safety and Cyber Risk Assessment

Capgemini has been selected to lead a consortium by the European Commission’s AI Office to support cyber threat modeling and risk assessment for advanced AI systems. This initiative is part of the EU’s AI Safety program, focusing on evaluating and mitigating cyber offense risks associated with General-Purpose AI (GPAI) models. The consortium, which includes Wavestone and the French public institution LNE, will help strengthen Europe’s capacity to assess, monitor, and mitigate systemic cyber offense risks from GPAI models. These models are rapidly developing autonomous and cyber-relevant capabilities, such as automated vulnerability discovery and large-scale cyber-attack orchestration. Capgemini will contribute its expertise in AI, cybersecurity, and compliance to develop evaluation tools, design technical compliance methodologies, and conduct technical assessments to identify systemic risks and enforce the AI Act’s obligations for GPAI providers.

CISA's 2030 Strategy Prioritizes AI and Advanced Technologies for Critical Infrastructure Security

CISA's 2030 Strategy Prioritizes AI and Advanced Technologies for Critical Infrastructure Security

The Cybersecurity and Infrastructure Security Agency (CISA) has released its 2030 strategy, outlining priorities for protecting the nation's critical infrastructure and federal networks. Advanced technologies, particularly Artificial Intelligence (AI), are central to this strategy. CISA plans to lead the assessment and deployment of transformative technologies like AI to enhance critical infrastructure operations. The agency will collaborate with government and industry partners to secure these innovative technologies. The strategy also addresses AI-enabled threats, with CISA working alongside critical infrastructure owners and operators to counter cyberthreats. Nick Andersen, CISA's acting director, emphasized a shift from passive defense to an aggressive, anticipatory stance to eliminate adversarial footholds in defense-critical infrastructure.

Operational Technology Cyber Coalition Urges 10-Year Reauthorization of CISA 2015 Amid AI-Driven Cyber Threats

Operational Technology Cyber Coalition Urges 10-Year Reauthorization of CISA 2015 Amid AI-Driven Cyber Threats

The Operational Technology Cyber Coalition (OTCC), joined by 34 other critical infrastructure organizations, has called on U.S. congressional leaders to include a ten-year reauthorization of the Cybersecurity Information Sharing Act of 2015 (CISA 2015) in the final Fiscal Year 2027 National Defense Authorization Act (NDAA). In a November 1 letter to the leaders of the House and Senate Armed Services Committees, the coalition supported Section 1706 of H.R. 8800, which would extend the law through 2035. CISA 2015 is set to expire on December 11, 2026. The law facilitates voluntary sharing of cyber threat indicators and defensive measures between private-sector entities and the federal government, offering crucial liability protections. The OTCC emphasized the law's importance for operational technology (OT) environments, where vulnerabilities in industrial control systems can disrupt essential services like electricity grids, drinking water systems, and manufacturing facilities. The coalition highlighted tha...