| | | | |  | | By Gregory Svirnovskiy, Daniel Payne, Ruth Reader and Erin Schumaker | | | | |  The Council of Europe wants America in the AI treaty, but that'll come at a price. | Getty Images | European countries negotiating a treaty on responsible use of artificial intelligence want U.S. backing — since America boasts the sector’s leading firms, including OpenAI, Google and Microsoft. But the price, as laid out by American diplomats, is language allowing the U.S. to exempt American companies from any new obligations the treaty imposes. That’s not going over well with some allies across the pond as well as activist groups that back tighter controls on AI. They say that any deal must bind both governments and corporations to be effective. The debate: The treaty, negotiated by the Council of Europe and its 46 member countries, would establish basic obligations to respect human dignity, the rule of law and democratic principles when AI is used. The U.S. is a non-voting observer at the council, but the Europeans crave its signature on the treaty because of the United States' importance in the development of AI. To get U.S. backing, American negotiators have demanded language allowing the U.S. to exempt firms, our POLITICO colleagues in Europe report. The idea is to protect U.S. national security interests. American activists at the Center for AI and Digital Policy, and Public Citizen, among other organizations, wrote to Secretary of State Antony Blinken to complain that an opt-out would gut the deal. “A hollowed-out treaty will provide little meaningful protections to individuals who will become increasingly subject to powerful AI systems prone to bias,” they wrote. What’s next? Partly under pressure from member countries like France and Germany, the EU is trying to meet Washington halfway in order to finalize the deal this week. EU member countries are now proposing to amend the draft treaty to allow all signatory countries to opt out of rules for companies.
| | | | JOIN US ON 3/21 FOR A TALK ON FINANCIAL LITERACY: Americans from all communities should be able to save, build wealth, and escape generational poverty, but doing so requires financial literacy. How can government and industry ensure access to digital financial tools to help all Americans achieve this? Join POLITICO on March 21 as we explore how Congress, regulators, financial institutions and nonprofits are working to improve financial literacy education for all. REGISTER HERE. | | | | | | | | | 
Rocky Mountain National Park, Colo. | Shawn Zeller/POLITICO | This is where we explore the ideas and innovators shaping health care. Say it ain't so, Joe. A Harvard researcher is examining a potential link between olive oil and cancer. Share any thoughts, news, tips and feedback with Carmen Paun at cpaun@politico.com, Daniel Payne at dpayne@politico.com, Ruth Reader at rreader@politico.com or Erin Schumaker at eschumaker@politico.com. Send tips securely through SecureDrop, Signal, Telegram or WhatsApp.
| | |  Ear infections are the bane of childhood for many. | Getty Images | Ear infections plague millions of young children. Untreated, the infections can lead to hearing loss and learning disabilities. AI could help prevent that, University of Pittsburgh researchers believe. How’s that? The researchers spent five years compiling a library of 1,151 videos featuring the eardrums of 635 children. They used the videos to train and test two AI tools to better recognize infection. They found it boosted the accurate diagnosis rate to more than 90 percent. That’s a big improvement over current methods. Doctors typically miss more than 1 in 4 cases. What’s next? By more accurately diagnosing the condition, the researchers hope to ensure kids get treated. At the same time, they hope to prevent overly cautious doctors from prescribing antibiotics when they’re not needed, an action that can reduce the drugs’ effectiveness among the general population over time.
| | | | DON’T MISS POLITICO’S HEALTH CARE SUMMIT: The stakes are high as America's health care community strives to meet the evolving needs of patients and practitioners, adopt new technologies and navigate skeptical public attitudes toward science. Join POLITICO’s annual Health Care Summit on March 13 where we will discuss the future of medicine, including the latest in health tech, new drugs and brain treatments, diagnostics, health equity, workforce strains and more. REGISTER HERE. | | | | | | | | |  Dermatology apps aren't ready for prime time, researchers found. | Getty Images | AI dermatology apps might be harming patients, according to a new analysis by researchers at UCLA, New York University, Stanford and others. They looked at dozens of apps that claim to detect skin cancer, track moles and recommend treatment for acne, among other conditions, and found a host of problems — from a lack of evidence and input from dermatologists to back-up diagnoses and treatment recommendations along with opaque algorithms and questionable data-privacy practices. How so? The analysis reviewed apps mostly aimed at consumers. None of the apps reviewed are FDA-approved, though only two had disclaimers saying so. Ten apps claimed diagnostic capabilities, though few offered evidence published in peer-reviewed journals. Because direct-to-consumer health tech is booming, the authors argue, regulations are needed to validate the technology and protect patients. Even so: The researchers offered hope for the future of such apps, saying proper safeguards could amplify their benefits and diminish their risks. That might already be happening in some instances: Certain consumer-facing apps have been classified in the EU as medical devices, offering hopes for future FDA review and quality assurances for users. | | | | Follow us on Twitter | | | | Follow us | | | | |
No comments:
Post a Comment