Trump, Xi Discuss AI Safety Amidst US-China Competition

President Donald Trump welcomed Chinese President Xi Jinping to the White House for a high-profile state visit focused on trade, Taiwan and artificial intelligence. Treasury Secretary Scott Bessent said the two countries agreed to a temporary trade deal. Xi pressed for a permanent halt to U.S. arms sales to Taiwan, while both leaders also addressed the growing competition and safety concerns surrounding AI development.

Xi Jinping stated that the U.S. and China share an obligation to keep AI under human control. Trump, however, ruled out new AI guardrails during the summit, saying he wants to leave regulation as it is and dismissing global safety efforts as a hoax. He pointed to the Department of Justice as the U.S. framework for oversight and rejected calls for a United Nations agreement on AI safety.

Tech industry leaders were well represented at the event. OpenAI CEO Sam Altman, Nvidia CEO Jensen Huang, Jeff Bezos, Sundar Pichai and Elon Musk were expected to attend the state dinner. Altman urged global cooperation on AI safeguards, while Huang argued that AI developers should not receive exemptions from existing antitrust or product-liability rules. This stance contrasted with a proposal from Anthropic CEO Dario Amodei, who suggested an antitrust waiver allowing AI labs to cooperate on safety.

Outside the summit, voices from the business world stressed the intensity of the AI race. Joshua Cooper Ramo, chairman and CEO of Sornay, said you must be in China to truly grasp the pace of AI innovation. Jason Rathje, a former DOD director and president of webAI, warned that China plans to win the AI race through scale and open-source model distribution. He urged the U.S. to rethink centralized data centers and build AI that people actively choose over Chinese alternatives.

Public opinion reveals a sharp disconnect with the administration's stance. A CNN poll found that 75% of Americans are worried about AI progress, and 72% believe the government should do more to regulate it. A separate survey by SSRS found 63% of respondents are concerned about AI's societal impact and 61% believe it will hurt the job market. Trump has repeatedly dismissed these concerns.

Several companies are moving forward on AI safety and security in the private sector. OpenAI plans to expand independent safety assessments into model training and evaluation, working with groups like METR and Redwood Research. RBC Capital Markets highlighted growing demand for Okta's identity platform as companies secure AI agents, and Stifel raised its Okta price target to $215 from $150. Cloud Range launched an AI Validation Range for security testing, and LatticeFlow AI introduced a managed risk assessment service for enterprises.

Product releases also continued across the industry. BottleCap AI released ThinkingCap-Qwen3.8-27B, a fine-tuned model that uses 37.2% fewer thinking tokens across 12 benchmarks. Teradata embedded WisdomAI agentic analytics into its Tera platform, aiming for over 95% accurate agentic business intelligence within a governed environment. The Disability Rights Education and Defense Fund launched a survey examining how AI tools affect legal services for disabled clients.

Key Takeaways

  • Trump welcomed Xi Jinping to the White House for a state visit covering trade, Taiwan and AI, with a temporary trade deal reportedly agreed upon.
  • Xi called for AI development to stay under human control, while Trump ruled out new AI guardrails and dismissed safety concerns as a hoax.
  • OpenAI CEO Sam Altman urged global AI cooperation, while Nvidia CEO Jensen Huang opposed antitrust exemptions for AI developers, contrasting with Anthropic CEO Dario Amodei's safety cooperation proposal.
  • Joshua Cooper Ramo and Jason Rathje stressed China's intense AI innovation and scale-based strategy, urging the U.S. to rethink centralized data centers and improve its AI offerings.
  • A CNN poll found 75% of Americans fear AI growth and 72% want more government regulation, contradicting the Trump administration's stance.
  • OpenAI plans to expand independent safety assessments into model training, working with METR and Redwood Research before models are finished.
  • Okta is seeing stronger demand for AI security and identity management, with Stifel raising its price target to $215 from $150.
  • Cloud Range launched an AI Validation Range for security testing, and LatticeFlow AI introduced a managed risk assessment service for enterprises.
  • BottleCap AI released ThinkingCap-Qwen3.8-27B, reducing thinking tokens by 37.2% across 12 benchmarks while maintaining 85.79% macro-average accuracy.
  • Teradata integrated WisdomAI agentic analytics into Tera, targeting over 95% accurate governed business intelligence, and DREDF launched a survey on AI impacts in legal services for disabled clients.

Trump and Xi meet at White House to discuss trade, Taiwan and AI

President Donald Trump welcomed Chinese President Xi Jinping to the White House for a state visit. The two leaders planned to discuss trade, Taiwan and artificial intelligence. Treasury Secretary Scott Bessent said the U.S. and China agreed to a temporary trade deal. Xi wants the U.S. to permanently halt arms sales to Taiwan. The visit also focused on the AI technology race and safety standards.

Sornay CEO says you must be in China to grasp AI innovation intensity

Joshua Cooper Ramo is chairman and CEO of Sornay. He spoke about the AI race between the U.S. and China on CNBC. He said you cannot understand the intensity of innovation in China unless you are there. He also shared his expectations for the meeting between President Donald Trump and Chinese leader Xi Jinping at the White House.

Xi says China and US must keep AI under human control

Chinese President Xi Jinping said the U.S. and China have an obligation to manage AI development. He spoke at the White House and said both nations should keep AI under human control. President Trump said he and Xi would discuss AI. OpenAI CEO Sam Altman urged global cooperation on AI safeguards. Trump has called concerns about AI dangers a hoax and wants the U.S. to lead the AI race.

Former DOD director urges better US AI strategy to counter China

Jason Rathje, president of the AI company webAI and a former DOD director, said the U.S. needs better AI to compete with China. He said China plans to win the AI race on scale. He urged the U.S. to rethink where AI lives and move away from centralized data centers. He noted that China's open-source AI models are spreading worldwide. He said American AI must be better so people choose it over Chinese alternatives.

Trump, Xi to face off over trade, Taiwan and AI at White House summit

President Donald Trump welcomed Chinese President Xi Jinping to the White House for a high-profile state visit. The two leaders were set to discuss trade, Taiwan and artificial intelligence. Treasury Secretary Scott Bessent said the U.S. and China agreed to a new trade deal. Xi wants the U.S. to halt arms sales to Taiwan. OpenAI CEO Sam Altman, Nvidia CEO Jensen Huang and Jeff Bezos were expected to attend the state dinner. China has also resisted U.S. pressure over its ties with Iran.

Trump plans AI talks with Xi but wants no changes

President Donald Trump said he will discuss artificial intelligence with Chinese President Xi Jinping at the White House. He said he wants to leave AI regulation exactly where it is. Trump has pushed back on global efforts to coordinate AI rules. He also dismissed concerns about AI as a hoax. Treasury Secretary Scott Bessent is working on related issues with Chinese officials.

Trump rules out AI guardrails during Xi summit

President Donald Trump said he does not expect any moves to put guardrails around artificial intelligence during his summit with Chinese President Xi Jinping. He posted that Super Intelligence will be a big topic but he wants to leave it as it is. He said the US guardrail is the Department of Justice. Trump also dismissed calls for an international agreement on AI safety at the United Nations.

Trump and Xi meet at White House with AI leaders

President Trump and Chinese President Xi Jinping met at the White House for several hours ahead of a state dinner. Tech leaders including Jeff Bezos, Sundar Pichai, Sam Altman and Elon Musk were set to attend. Xi stressed that AI development must stay under human control. Xi also called for more cooperation, direct flights and cultural exchanges between the two countries.

RBC says AI security could boost Okta demand

RBC Capital Markets said Okta could see stronger demand for its identity platform. Companies are adopting tools to secure artificial intelligence and machine learning systems. This creates a growing need for identity and access management solutions. Okta's platform helps organizations manage user identities and protect against cyber attacks. The report said Okta's stock could keep rising as demand grows.

RBC highlights Okta's AI agent security growth

RBC Capital Markets said Okta could see stronger demand as companies secure artificial intelligence agents. The growing use of AI agents in many industries could drive interest in Okta's identity and access management solutions. Companies also need to protect their AI agents from threats. This could lead to more adoption of Okta's platform.

Stifel raises Okta price target to $215 on AI security growth

Stifel raised its price target for Okta Holdings to $215 from $150. The increase reflects Okta's push into AI security and growing demand for cloud-based security services. Okta's stock has gained about 20% in the past year. The new target is higher than the average analyst estimate of $180.

Most Americans fear AI and want more government regulation

A CNN poll found that 72% of Americans believe the government should do more to regulate AI. 63% of respondents said they are concerned about AI's impact on society. 61% believe AI will negatively affect the job market. The poll was conducted by SSRS from September 16 to 20 among 1,003 adults.

CNN poll shows 75% of Americans fear AI growth

A CNN poll released Thursday found that three quarters of Americans are worried about AI progress. 71% said the federal government is not doing enough to regulate AI. President Trump called concerns about AI regulation a hoax in a Truth Social post. The poll surveyed 1,206 respondents from September 16 to 17.

OpenAI expands outside safety reviews into model training

OpenAI plans to expand independent safety assessments into model training and evaluation. The company is talking with AI research groups METR and Redwood Research as potential assessors. OpenAI wants outside evaluators involved before models are finished, not just before launch. Assessments will include strong independence mechanisms and scientific rigor.

BottleCap AI releases ThinkingCap-Qwen3.8-27B with fewer tokens

BottleCap AI released ThinkingCap-Qwen3.8-27B, a fine-tuned version of Qwen3.8-27B. The model uses 37.2% fewer thinking tokens on average across 12 benchmarks. Macro-average accuracy drops slightly from 86.65% to 85.79%. It is available on vLLM and SGLang with multiple build options including FP8 and GGUF.

Teradata embeds WisdomAI agentic analytics in Tera platform

Teradata integrated WisdomAI's deterministic Agentic Analytics platform into Tera within its Autonomous Knowledge Platform. The system supports governed natural language analytics, app creation from a single prompt, and proactive AI agents across structured and unstructured data. Broader availability on Teradata Cloud is expected after testing in 2026. The integration aims to achieve over 95% accurate agentic BI while keeping AI execution inside Teradata's governed environment.

Cloud Range launches AI Validation Range for security testing

Cloud Range introduced the AI Validation Range, a contained testing environment where security operations centers can benchmark AI agents against human defenders. The platform recreates realistic enterprise settings and uses automated adversary attack emulations to test AI under adversarial and unexpected conditions. It helps teams determine which security roles suit AI, which need human oversight, and where combining both works best. The tool is built on the Cloud Range cyber range platform and supports the Cloud Range AI Readiness Framework.

LatticeFlow AI launches managed risk assessment service

LatticeFlow AI announced the LatticeFlow AI Risk Center, a governance service that helps enterprises assess AI risk and scale AI systems with confidence. The service extends enterprise AI teams by continuously evaluating AI systems for security, performance, and other critical risks using LatticeFlow AI technology and expertise. Customers receive regular risk reports with actionable mitigation guidance and risk review sessions with stakeholders. The service is designed as an ongoing operation rather than a one-time assessment.

Nvidia CEO opposes liability exemptions for AI developers

Nvidia CEO Jensen Huang stated that AI developers should not be exempt from existing antitrust or product-liability rules. His comments came during an interview with Ezra Klein and contrasted with a proposal from Anthropic CEO Dario Amodei for an antitrust waiver allowing AI labs to cooperate on safety. Huang argued that developers remain responsible for testing models before public release. He also said regulators may need additional rules for specific applications like autonomous vehicles.

DREDF surveys legal services clients on AI impacts

The Disability Rights Education and Defense Fund is conducting an online survey about AI and legal services for disabled clients. The survey targets attorneys, legal fellows, paralegals, law students, and legal intake staff. It aims to learn whether AI tools have affected people with disabilities in areas like healthcare, employment, housing, and education. DREDF also seeks to understand legal services providers' interest in AI training related to disability issues.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI Artificial Intelligence AI Safety AI Regulation AI Race AI Innovation AI Security AI Development AI Cooperation AI Concerns AI Polls AI Products AI Releases AI Companies AI Industry AI Progress AI Impact AI Job Market AI Governance AI Business Intelligence

Comments

Loading...