Harvard T.H. Chan School of Public Health wrapped up a four-part pilot series this fall on ethical AI use, drawing more than 100 community members. Sessions covered generative AI tools like Claude and ChatGPT, highlighting their ability to summarize notes and generate text while warning that they can produce false information. The series emphasized verifying AI outputs and keeping human thinking at the center of research and learning.
The U.S. government ordered federal civilian agencies to remove DeepSeek AI products from their systems and stop purchasing them. The ban targets Hangzhou DeepSeek Artificial Intelligence Co., Ltd and its R1 and V3 models. The Department of Homeland Security flagged concerns that user data could be accessed by Chinese authorities. The order applies to federal agencies and contractors only, not to businesses or the general public.
Experts argue that businesses need artificial specific intelligence rather than artificial general intelligence. Models from OpenAI and Anthropic, while powerful, are frozen at training and cannot absorb a company's unique knowledge. The recommended approach involves defining clear success metrics, selecting the right pilot project, and using context engineering to guide AI behavior. The authors stress that computing power alone cannot compensate for a lack of business-specific understanding.
At the third annual Innovations in Artificial Intelligence Conference held at UAMS on Oct. 8-9, more than 150 attendees gathered to discuss ethical AI in healthcare. A key concern was AI scribes: only 25% of tools have privacy policies in place. Joseph Sanford reported that AI scribe use in Arkansas clinics jumped from 6.8% to over 12% of patients in just five months. Panels stressed keeping humans in the loop and building critical thinking alongside AI tools.
Microsoft partnered with consumer healthcare company Opella to build an agentic AI platform for verifying health content at scale. The platform will use AI to analyze and ensure the accuracy of health information, aiming to address the growing need for trustworthy content online. The tool is designed to help healthcare professionals, researchers, and the public access reliable information.
With over 1600 AI medical devices now authorized, pharmacists are emerging as the critical link between these tools and patients. A survey of 122 pharmacists found 61% had moderate familiarity with AI tools, yet most use them only for administrative tasks. Concerns about accuracy, lack of clinical judgment, and data privacy risks remain widespread. Rita King of MethodSense says pharmacists must carefully check device instructions to prevent patient misunderstanding and errors.
Rockwell Automation launched FactoryTalk Conductor, an AI-native orchestration platform that coordinates multiple AI agents and workflows for industrial automation. Engineers can start tasks using natural language, and the platform connects tools like Studio 5000 Logix Designer, Logix Echo, Emulate3D, and FactoryTalk Optix. It supports enterprise AI governance and can work with locally hosted open-weight models to reduce token costs. The platform was showcased at Automation Fair 2026.
Oil prices rose and Asian equities fell as focus shifted back to Middle East geopolitical developments, easing earlier AI market jitters. President Trump stated that the U.S. will not attack, according to the report. The shift in market attention highlights how AI-related sentiment can quickly give way to traditional geopolitical drivers.
Korea's Fair Trade Commission released draft Guidelines on the Labeling and Advertising of Artificial Intelligence for public comment from October 22 through November 11. The rules target false claims, deception, unfair comparisons, and disparagement in AI marketing. A consumer survey found 57.9% of people would pay about 21% more for AI products, and the KFTC included 16 simulated cases to help businesses and consumers navigate the new standards.
Key Takeaways
- Harvard's four-part AI ethics series trained over 100 community members to verify outputs from tools like Claude and ChatGPT and keep human thinking central.
- The U.S. banned federal agencies from buying or using DeepSeek R1 and V3 models over data access concerns tied to Chinese authorities.
- Experts say businesses should pursue artificial specific intelligence, using context engineering and clear metrics, since OpenAI and Anthropic models cannot learn company-specific knowledge.
- At UAMS's AI conference, AI scribe use in Arkansas clinics rose from 6.8% to over 12% of patients in five months, with only 25% of tools having privacy policies.
- Microsoft and Opella are building an agentic AI platform to verify health content at scale for professionals and the public.
- Over 1600 AI medical devices are now authorized, but a survey of 122 pharmacists found most use AI only for administrative tasks and worry about accuracy and privacy.
- Rockwell Automation launched FactoryTalk Conductor, an AI-native platform that orchestrates multiple agents and connects tools like Studio 5000 Logix Designer and Emulate3D.
- Oil prices rose and Asian equities fell as Middle East tensions returned and AI market jitters eased; Trump stated the U.S. will not attack.
- Korea's Fair Trade Commission released draft AI marketing guidelines covering false claims and unfair comparisons, open for public comment through November 11.
- A Korean consumer survey found 57.9% of people would pay about 21% more for products marketed as AI-powered.
Harvard Chan School explores ethical AI use in learning and research
Harvard T.H. Chan School of Public Health held a four-part pilot series this fall to teach ethical AI use. More than 100 community members attended sessions on generative AI, research tools, data analysis, and workplace skills. Librarian Carrie Wade explained that LLMs like Claude and ChatGPT can summarize notes and generate text but may produce false information. The series stressed verifying all AI outputs and using tools responsibly while keeping human thinking central.
US bans federal agencies from buying DeepSeek AI products
The US ordered federal civilian agencies to remove DeepSeek AI products from their systems and stop purchasing them. The ban targets Hangzhou DeepSeek Artificial Intelligence Co., Ltd and its R1 and V3 AI models. The Department of Homeland Security raised concerns that user data could be accessed by Chinese authorities. The order applies only to federal agencies and contractors, not to businesses or the general public.
Enterprises need specific AI, not general intelligence, experts say
A new opinion piece argues that businesses need artificial specific intelligence rather than artificial general intelligence. Models from OpenAI and Anthropic are powerful but frozen at training and cannot learn a company's unique knowledge. The article suggests three steps for success: define your own success metrics, choose the right pilot project, and use context engineering to guide AI. The author says no amount of computing power can fix the lack of business-specific understanding.
UAMS hosts third annual AI conference on ethical healthcare use
The third annual Innovations in Artificial Intelligence Conference at UAMS drew more than 150 attendees on Oct. 8-9. Speakers discussed ethical AI use in healthcare, including concerns about AI scribes, with only 25% having privacy policies. Joseph Sanford noted AI scribe use in Arkansas clinics jumped from 6.8% to over 12% of patients in five months. The panel stressed keeping humans in the loop and developing critical thinking and curiosity alongside AI tools.
Microsoft and Opella build AI health content verification platform
Microsoft and consumer healthcare company Opella announced a partnership to create an agentic AI platform for verifying health content at scale. The platform will use AI to analyze and ensure the accuracy of health information. The collaboration aims to address the growing need for trustworthy health information online. The tool will help healthcare professionals, researchers, and the public access reliable content.
Pharmacists play key role in safe AI medical device use
Over 1600 AI medical devices are now authorized, and pharmacists are the main link between these tools and patients. Rita King of MethodSense says pharmacists must check device instructions to prevent patient misunderstanding and errors. A survey of 122 pharmacists found 61% had moderate familiarity with AI tools, but most used them only for administrative tasks. Many pharmacists worry about accuracy, lack of clinical judgment, and data privacy risks with these devices.
Rockwell Automation launches AI orchestration platform FactoryTalk Conductor
Rockwell Automation introduced FactoryTalk Conductor, an AI-native platform that coordinates multiple AI agents and workflows for industrial automation. The platform lets engineers start engineering tasks using natural language and connects tools like Studio 5000 Logix Designer, Logix Echo, Emulate3D, and FactoryTalk Optix. It supports enterprise AI governance and can work with locally hosted open-weight models to reduce token costs. The platform was showcased at Automation Fair 2026 and aims to speed up engineering work across the automation lifecycle.
Oil prices rise as Middle East tensions return and AI market worries ease
Oil prices rose and Asian equities fell as attention shifted back to Middle East geopolitical developments. Concerns about a possible re-escalation of the Middle East conflict pushed oil higher early Thursday. The shift came as artificial intelligence market jitters eased. President Trump has stated that the U.S. will not attack, according to the report.
Korea develops guidelines to fight AI washing in product marketing
The Korea Fair Trade Commission is creating standards to stop companies from exaggerating or faking AI capabilities in their products. A draft of the Guidelines on the Labeling and Advertising of Artificial Intelligence was released for public comment from October 22 through November 11. The guidelines cover false claims, deception, unfair comparisons, and disparagement in AI marketing. A consumer survey found 57.9% of people would pay about 21% more for AI products, and the KFTC included 16 simulated cases to help businesses and consumers understand the rules.
Sources
- AI’s role in learning, research, and career exploration—using tools thoughtfully, ethically | Harvard T.H. Chan School of Public Health
- US bans federal agencies from buying DeepSeek AI products
- Enterprises don’t need artificial general intelligence. They need artificial specific intelligence
- Third Annual AI Conference at UAMS Focuses on Ethical, Responsible Use
- Microsoft, Opella partner on agentic AI content verification platform
- Pharmacist Roles, Risks in Guiding AI-Enabled Medical Device Use
- Rockwell Automation Introduces FactoryTalk Conductor, an AI-Native Orchestration Platform for Industrial Automation
- Oil Rises as Geopolitics Back in Focus After AI Jitters Ease
- No AI Technology, Yet an AI Product? ... Comments Sought on Guidelines to Crack Down on 'AI Washing'
Comments
Please log in to post a comment.