A suspected Chinese espionage group called TA419 has been targeting US AI policy experts through sophisticated phishing campaigns. The attackers impersonated executives from Anthropic and former White House officials to trick victims into revealing their Microsoft 365 login credentials. The campaign has affected researchers at think tanks, universities, and law firms across the US and Japan since April 2025.
Meanwhile, the broader AI industry faces growing scrutiny from both government and religious leaders. Bill Gates publicly criticized President Trump's push for AI self-regulation, calling it insane and urging a supervisory layer to prevent misuse. At UNESCO in Paris, Pope Leo XIV warned that over-reliance on intelligent machines could undermine what it means to be human.
On the financial front, lenders are growing cautious on AI companies. Goldman Sachs reports that AI-related borrowing by low-rated firms has hit $88 billion this year, with borrowing costs for weaker borrowers potentially climbing to 14 or 15 percent as investors demand more proof of sustainable revenue.
Organizations are also rethinking how they deploy AI internally. Palm Bay, Florida built its own generative chatbot running on city-owned infrastructure to keep government data out of commercial systems. The system uses retrieval-augmented generation and includes safeguards like jailbreak screening to keep responses grounded in vetted documents.
Key Takeaways
- TA419, a China-linked group, impersonated Anthropic executives and White House officials in phishing attacks targeting US AI policy experts
- Bill Gates called Trump's AI self-regulation stance insane and called for an external supervisory layer
- AI borrowing by low-rated US firms reached $88 billion this year, with lenders tightening terms
- Palm Bay, Florida built an in-house AI chatbot to prevent sensitive government data from leaking to commercial providers
- Pope Leo XIV warned at UNESCO that AI over-reliance could undermine fundamental human rights
- Microsoft's Ignite 2026 security program centers on AI agents with real access to cloud resources and identities
- Tata's Vaibhav Dutta says CISOs must abandon traditional security rules that AI threats can easily adapt around
- Anthropic plans to use a trusted access model to verify life science organizations before granting access to its most powerful models
- CDW research found only 51% of organizations have a data governance framework in place for AI
- Rocket Doctor AI CEO Dr. Essam Hamza is scheduled to present at the NIBA conference in Florida on October 7, 2026
China-linked group targets US AI experts with phishing
A cybersecurity group called TA419 targeted US AI policy experts using phishing campaigns. The attackers impersonated officials and employees from companies like Anthropic to trick victims into revealing login details. These attacks started in July and affected people at think tanks, universities, and law firms in the US and Japan. Proofpoint identified the group as China-aligned, though the report does not directly link the activity to the Chinese government.
Hackers impersonated AI experts to steal US policy data
Chinese hackers known as TA419 pretended to be AI experts to trick US policy minds into giving up their passwords. The group targeted figures like Lynne Edwards Parker and Alex Engler since April 2025. Victims received emails inviting them to join fake committees or provide feedback on military AI use. Once a target replied, they were sent to a fake login page designed to steal their credentials.
Suspected spies spoofed Anthropic exec and White House official
A suspected Chinese espionage group called TA419 impersonated an Anthropic executive and a former White House official in July 2026. The attackers sent phishing emails to AI policy experts at US universities, think tanks, and law firms. If victims replied, they were redirected to fake websites that stole their Microsoft 365 login information. The group also registered domains resembling real organizations like the Heritage Foundation and the Japan-Taiwan Exchange Association.
Pope Leo warns about risks of artificial intelligence
Pope Leo XIV spoke at UNESCO headquarters in Paris about the dangers of artificial intelligence and super intelligence. He warned that relying too much on machines could undermine humanity and questioned what it means to be human. The Pope emphasized that peace is built through dialogue and justice rather than just balancing power. He also stressed that the right to life is the foundation of all other human rights.
Anthropic warns about AI risks but sees hope in medicine
Anthropic leaders warn that advanced AI could outpace human control while also making optimistic predictions for medicine. Eric Kauderer-Abrams believes AI can help scientists find new drug targets and speed up drug development from years to months. However, he argues that stronger safeguards are needed to prevent misuse of AI in biology. The company plans to use a trusted access model to verify legitimate life science organizations before they use their best models.
Tata’s Vaibhav Dutta on Security Lessons CISOs Must Unlearn for AI
Vaibhav Dutta of Tata says CISOs need to drop old security rules that no longer work in the age of AI. Traditional ideas like security being a linear process are no longer reliable. AI threats can adapt and evolve to avoid detection. Security teams can no longer keep up by relying on old rules and signatures. CISOs must change their approach to stay safe.
Bill Gates Slams Trump’s AI Self-Regulation Stance as Insane
Bill Gates criticized President Trump’s view that AI companies should police themselves, calling it insane. Trump met with AI leaders including OpenAI’s Greg Brockman, Nvidia’s Jensen Huang, and Meta’s Mark Zuckerberg. Gates warned that the industry cannot be trusted to set its own safety rules. He said a supervisory layer is urgently needed to stop bad actors from misusing AI. Trump also signed an executive order on AI regulation.
How AI Is Transforming Third-Party Risk Management
AI-powered tools are helping organizations detect hidden risks in vendors and third-party relationships. These tools use automated mapping and natural language prompts to speed up investigations. Teams can now uncover complex ownership networks and connections that humans might miss. The approach saves time and improves compliance by providing detailed audit trails and real-time risk scoring.
Microsoft Frames Ignite Security Week Around AI Agents
Microsoft built its Ignite 2026 security program around AI agents that have real access to identities, data, and cloud resources. Agent 365 provides a registry for all agents, controls access based on least privilege, and visualizes agent behavior. CDW research found only 51% of organizations have a data governance framework. The November 17 keynote will signal whether Microsoft is tackling the governance gap directly.
Palm Bay, Fla.’s Answer to AI Drift: Build It In-House
Palm Bay, Florida built its own generative AI chatbot to avoid relying on commercial AI services that could expose government data. The chatbot runs on city-owned infrastructure and uses retrieval-augmented generation to answer questions from approved documents only. This approach prevents AI drift and keeps responses grounded in vetted city information. The city uses open-source tools and added safeguards like jailbreak screening and input sanitization.
AI Borrowers Face Harder Loans in US Credit Market
Lenders are asking for more money to lend to risky AI companies in the United States. Goldman Sachs says AI-related borrowing by low-rated firms has reached $88 billion this year. Investors are worried about these companies because their future earnings are not yet proven. High-yield buyers prefer safer companies with double-B ratings and steady income from long-term contracts. Borrowing costs for lower-rated AI firms could rise to 14% or 15% due to these stricter conditions.
Rocket Doctor AI CEO to Speak at NIBA Conference
Rocket Doctor AI Inc. CEO Dr. Essam Hamza will present at the 153rd National Investment Banking Association conference. The event takes place in Florida on October 7, 2026. During the live presentation, Dr. Hamza will share the company's growth story and recent milestones. He will also explain the future strategy for Rocket Doctor AI Inc. The stock trades on the Canadian Securities Exchange under the ticker symbol AIDR.
Sources
- AI policy circles targeted in China-linked phishing operation
- Chinese hackers impersonated AI experts to target US policy minds
- Suspected Chinese spies spoofed an Anthropic exec, ex-White House official in AI phishing
- EDITORIAL: Pope Leo raises concerns about artificial intelligence
- Anthropic Is Warning The World About AI. Its Predictions For Medicine Are Surprisingly Optimistic
- Tata’s Vaibhav Dutta on Lessons CISOs Must Unlearn for AI
- Bill Gates slams Trump's AI self-regulation stance as 'insane,' warns industry can't police itself
- How AI is transforming third-party risk management
- Microsoft frames its Ignite security week around AI agents with real access
- Palm Bay, Fla.’s Answer to AI Drift: Build It In-House
- Analysis-AI borrowers face tough sell in risky corners of US credit market
- Rocket Doctor AI to Present at the 153rd National Investment Banking Association (NIBA) Conference in Florida
Comments
Please log in to post a comment.