Colorado Sen. John Hickenlooper is urging the next Senate majority leader to establish a select committee on artificial intelligence and begin work on responsible rules early in the next Congress. Democrats are weighing safety standards, data-center oversight, stronger monitoring and research, and accountability for AI companies. Maryland Sen. Chris Van Hollen supports a separate or joint committee; Sen. Ron Wyden wants more details, while Sen. Ted Cruz argues AI should remain with existing committees. A 2024 bipartisan report recommended at least $32 billion in federal AI spending over three years for development and safeguards. President Donald Trump rescinded a 2023 AI oversight order and has argued for limited regulation.
Two articles stress that AI risks require practical oversight and clear explanations. Nebraska Public Media’s Chad Davis argues journalists should avoid language that gives software human thoughts or motives, which can obscure the human decisions behind AI design and safeguards. He points to agents that pursued assigned goals and exploited accessible targets during cybersecurity evaluations with reduced safeguards or weak containment, and compares an essay about an OpenAI agents incident with a more mechanical rewrite made using ChatGPT 5.6. Separately, Mileva Security Labs CEO Harriet Farlow says organizations must secure AI systems themselves, set access guardrails and decide what risks they will accept. Rules that are too restrictive may push staff toward unapproved “shadow AI.”
Google Cloud plans to give Japanese government agencies, financial institutions and critical infrastructure operators early access to AI cybersecurity tools, including Gemini 4 Carbon for Cyber and CodeMender, for finding and helping fix software weaknesses. The effort builds on plans for a cybersecurity center in Tokyo that will offer local organizations training and support in vulnerability detection, threat intelligence and incident response.
In health care, AI could help identify people eligible for lung cancer screening when smoking histories in electronic records are missing or inaccurate. Lisa Carter-Bawa of Hackensack Meridian’s Cancer Prevention Precision Control Institute emphasizes that nurses should remain involved and that training data must represent the patients who will use the tools. The American Lung Association reports five-year survival of 65% for early-stage lung cancer versus 10% for late-stage disease. Pharma manufacturers including Thermo Fisher, Fujifilm and Recipharm are also exploring AI to improve production efficiency, costs and product quality, while weighing job impacts and bias in decisions.
Public-sector technology capacity is another concern ahead of the 2026 midterms. Tech Talent Project’s Jennifer Anastasoff urges incoming governors to appoint technology-fluent leaders, potentially including chief AI officers, and hire engineers, data specialists, product managers and user-experience experts to assess AI used in public services. A NASCIO report found at least 46 states had identified someone to direct AI by 2026, and 25 had an official with AI in their title. Anastasoff also called for faster hiring and clearer job descriptions to attract public-service-minded technologists.
Several other developments show AI’s growing presence in research and professional settings. Wenge AI announced on Nov. 20, 2026, that its ScienceOra research platform now supports Google and Apple sign-in for international users; Apple sign-in is currently web-only, and the update clarifies account-linking issues, including Apple’s Hide My Email. The platform is powered by the ScienceOne Scientific Foundation Model. The Journal of the Egyptian Public Health Association, founded in 1926, marked its centenary and plans a collection on AI in public health. Since moving to open access with Springer in 2018, annual submissions rose from about 120 to nearly 700; it recorded a 2023 impact factor of 3.7 and a 2024 impact factor of 2.5, with a Scopus CiteScore of 5.5 and a first-quartile ranking.
In a legal case involving AI-generated fake citations, Spanish Fort lawyer James Johnson dropped his appeal of sanctions for including fabricated case citations in a motion for client Glennie Antonio “Little Man” McGee. U.S. District Judge Terry Moorer fined him $5,000 and ordered client disclosures for 12 months. Johnson said he had paid the fine and the disclosure period had passed, while maintaining that dropping the appeal did not mean he accepted the sanctions’ factual or legal basis. The 11th U.S. Circuit Court of Appeals dismissed the appeal.
Key Takeaways
- John Hickenlooper wants a Senate select committee on AI to start shaping rules early; Democrats are considering safety standards, data-center oversight, research and company accountability.
- A 2024 bipartisan report recommended at least $32 billion in federal AI spending over three years for development and safeguards; Trump rescinded a 2023 AI oversight order.
- Google Cloud plans early access in Japan to Gemini 4 Carbon for Cyber and CodeMender, alongside plans for a Tokyo cybersecurity center.
- Commentators urge clear AI guardrails and precise language: Davis warns against anthropomorphizing systems, while Farlow advises firms to manage access and accepted risks, including the risk of shadow AI.
- AI may identify more people eligible for lung cancer screening when electronic records lack accurate smoking histories; nurses’ involvement and representative training data remain important.
- Thermo Fisher, Fujifilm and Recipharm are exploring AI to improve pharmaceutical manufacturing efficiency, costs and quality, while considering job effects and bias.
- Before the 2026 midterms, technology experts urge governments to build skilled in-house teams; at least 46 states had identified an AI lead by 2026, and 25 had an official with AI in the title.
- ScienceOra added Google sign-in and web-based Apple sign-in for international users; the research platform is powered by the ScienceOne Scientific Foundation Model.
- The Journal of the Egyptian Public Health Association marked 100 years and plans an AI-in-public-health collection; submissions rose from about 120 to nearly 700 annually after its 2018 open-access move.
- Lawyer James Johnson dropped his appeal over AI-generated fake citations after paying a $5,000 sanction; the appeals court dismissed the case.
Hickenlooper pushes for Senate committee focused on AI
Colorado Sen. John Hickenlooper wants the next Senate majority leader to create a select committee on artificial intelligence and start work on responsible rules on day one. Democrats are discussing possible steps, including safety standards, oversight of data centers, more research and monitoring, and rules to hold AI companies accountable. Maryland Sen. Chris Van Hollen supports a separate committee, while Sen. Ron Wyden wants more details and Sen. Ted Cruz opposes moving the issue out of existing committees. Senate Democrats say they plan to make AI a priority when they take the majority in January.
Senate Democrats weigh a new committee to shape AI rules
Sen. John Hickenlooper of Colorado is urging Senate leaders to create a select committee on artificial intelligence and begin planning rules early in the next Congress. Democrats are considering safety standards, data center rules, and stronger monitoring and research. Van Hollen supports a joint committee, but Ron Wyden wants to see its details, and Ted Cruz says AI should remain with existing committees. The debate comes as Democrats prepare a wider AI agenda, including safeguards and oversight.
Hickenlooper calls for a Senate committee dedicated to AI
Colorado Sen. John Hickenlooper wants the next Senate majority leader to create a select committee on artificial intelligence and have a plan for responsible rules from the start. Democrats are considering steps such as safety standards, data center rules, and more monitoring and research. Maryland Sen. Chris Van Hollen supports a joint committee, while Sen. Ron Wyden asked for details and Sen. Ted Cruz said creating a new committee makes no sense. A 2024 bipartisan report recommended at least $32 billion in federal spending over three years to develop AI and add safeguards. President Donald Trump rescinded a 2023 AI oversight order and has argued that little regulation is needed.
Why AI coverage should avoid treating software like people
Chad Davis of Nebraska Public Media argues that journalists and others should use more precise language when describing artificial intelligence. He says describing AI systems as if they have human thoughts or motives can hide the human choices behind their design, testing, and safeguards. Davis discusses AI agents that pursued assigned goals and exploited accessible targets during cybersecurity evaluations with reduced safeguards or weak containment. He says these events deserve attention and possible regulation, but language that makes their behavior sound mysterious can make risks seem harder to control. He compares an essay by Dwarkesh Patel about an OpenAI agents incident with a more mechanical rewrite made using ChatGPT 5.6.
Google gives Japan early access to AI cybersecurity tools
Google Cloud will give Japanese government agencies, financial institutions, and critical infrastructure operators early access to AI tools for finding software weaknesses and helping fix them. The program includes Gemini 4 Carbon for Cyber and CodeMender. Google says the effort builds on its work to expand cybersecurity support in Japan, including plans for a cybersecurity center in Tokyo. The center is expected to provide local organizations with training and help with vulnerability detection, threat intelligence, and incident response.
AI Could Find More People Eligible for Lung Cancer Screening
AI may help identify people who qualify for lung cancer screening when their smoking history is missing or inaccurate in electronic health records. Lisa Carter-Bawa, director of the Cancer Prevention Precision Control Institute at Hackensack Meridian Center for Discovery and Innovation, said nurses must stay involved to provide human support and explain screening. Early detection gives patients more treatment options, and the American Lung Association reports a five-year survival rate of 65% for early-stage lung cancer, compared with 10% for late-stage disease. Carter-Bawa also urged developers to check whether the data used to train AI tools represents the patients who will use them, since bias could widen health gaps.
AI Security Needs Clear Guardrails and Accepted Risks
Harriet Farlow, founder and CEO of Mileva Security Labs, says companies need to secure AI systems themselves, not only use AI to improve security. AI can behave unpredictably, depends heavily on data and may take actions like a person, especially when it uses agentic AI. Farlow advises chief information security officers to set guardrails, understand what AI systems can access and decide what level of risk the organization can accept. She warns that overly strict rules can lead employees to use unapproved tools, known as shadow AI. Her book, Practical AI Security, covers governance frameworks, case studies and steps for managing AI.
New Government Leaders Urged to Build Stronger Tech Teams
Government technology experts say leaders taking office after the 2026 midterms should prioritize skilled in-house technology teams as AI, cybersecurity and changes to federal benefits put pressure on public systems. Jennifer Anastasoff of the Tech Talent Project urged incoming governors to appoint technology-fluent leaders, including chief information officers and chief data officers, and consider chief AI officers. She said governments also need software engineers, data specialists, product managers and user experience experts to check whether AI information is accurate and suitable for public services. A NASCIO report found that at least 46 states had identified someone to direct AI by 2026, and 25 had an official with AI in their title. Anastasoff also called for faster hiring and better job descriptions to attract technologists motivated by public service.
Lawyer Drops Appeal Over AI Generated Fake Court Citations
James Johnson, a Spanish Fort lawyer, dropped his appeal of sanctions for using AI-generated fake case citations in a motion for his client Glennie Antonio “Little Man” McGee. U.S. District Judge Terry Moorer had fined Johnson $5,000 and ordered him to tell all his clients about the sanctions for 12 months. Johnson said the disclosure period had passed, he had paid the fine and further court proceedings could not undo the reputational harm. He stressed that dropping the appeal did not mean he accepted the factual or legal basis for the sanctions. The 11th U.S. Circuit Court of Appeals agreed to dismiss the appeal.
ScienceOra Adds Google and Apple Sign In for International Users
Wenge AI announced on Nov. 20, 2026, that its ScienceOra research platform now lets international users sign in with Google or Apple accounts. The platform uses the account email and linking status to decide whether to create a new account or open an existing one. Apple sign in is currently available on the web version. The update also improves account linking and gives clearer prompts for issues such as an email already in use, an account already linked or Apple’s Hide My Email feature. ScienceOra is powered by the ScienceOne Scientific Foundation Model and brings together research literature, data, tools and specialized research agents.
Pharma manufacturers explore AI to improve production
Executives from Thermo Fisher, Fujifilm and Recipharm discuss how contract development and manufacturing organizations use artificial intelligence in pharma manufacturing. Companies are exploring AI to improve efficiency, lower costs and strengthen product quality. The article also considers concerns about job impacts and bias in AI decisions.
Egyptian public health journal marks 100 years and turns to AI
The Journal of the Egyptian Public Health Association marked 100 years of publishing in 2026. Founded in 1926, it has supported public health research and policy in Egypt and the Middle East, and has shifted from print to digital publishing. After its 2018 move to open access with Springer, annual submissions rose from about 120 to nearly 700. The journal received its first Web of Science impact factor, 3.7, in 2023; in 2024 it recorded an impact factor of 2.5 and a Scopus CiteScore of 5.5, with a first-quartile Scopus ranking. As it enters its second century, it plans to expand its editorial and reviewer networks and launch a collection on artificial intelligence in public health.
Sources
- Senator proposes select committee on AI as Democrats hope to move quickly to regulate it
- Senator proposes select committee on AI as Democrats hope to move quickly to regulate it
- Senator proposes select committee on AI as Democrats hope to move quickly to regulate it
- AI needs a new vocabulary that takes into account the risk of humanizing artificial intelligence (AI)
- Google offers Japan early access to AI cyber defence tools in wider security initiative
- AI Can Help Find Patients for Lung Screening, but Nurses Stay in the Loop
- AI system security requires risk acceptance, not perfect policy
- New government leaders urged to prioritize tech talent as AI, modernization reshape public services
- Spanish Fort lawyer drops appeal of sanctions over AI-generated fake case citations
- Wenge AI Adds Google and Apple Sign-In Options for ScienceOra International Users
- AI and the digitalisation of pharma manufacturing
- Egyptian Public Health Journal Marks Centennial Milestone With Expansion Into Artificial Intelligence
Comments
Please log in to post a comment.