April 30, 2025

news   Read News
AI Breakthroughs in Healthcare, Cybersecurity, and Education

Recent developments in artificial intelligence have shown significant promise in various fields, including healthcare, cybersecurity, and education. Researchers at Uppsala University have developed an AI model that can diagnose multiple sclerosis with 90% accuracy, while another model can detect the progression of the disease earlier than traditional clinical diagnosis. The University of South Florida has launched a new college focused on AI, cybersecurity, and computing to address the growing demand for tech talent. Additionally, AI is being used to preserve fragments of human cognition and detect threats in security systems. However, concerns have been raised about the ethics of AI, including bias and the potential risks of error-prone AI-generated content. The US government has also urged the EU to delay its AI regulations, citing concerns about undue burdens on developers and stifling innovation. Meanwhile, a conservative activist is suing Meta over false responses generated by its AI chatbot, highlighting the need for tech companies to take responsibility for AI-generated content.

Key Takeaways

  • AI can diagnose multiple sclerosis with 90% accuracy and detect disease progression earlier than traditional clinical diagnosis.
  • The University of South Florida has launched a new college focused on AI, cybersecurity, and computing.
  • AI is being used to preserve fragments of human cognition and detect threats in security systems.
  • Concerns have been raised about the ethics of AI, including bias and error-prone AI-generated content.
  • The US government has urged the EU to delay its AI regulations, citing concerns about undue burdens on developers.
  • A conservative activist is suing Meta over false responses generated by its AI chatbot.
  • AI can help reduce the number of failures in service dog training and improve the selection process.
  • The RSA Conference will feature talks on the use of AI technologies in threat detection and automated vulnerability management.
  • Apollo.io has appointed new executives to accelerate AI-powered go-to-market innovation.
  • The National Science Foundation canceled a grant for a research project to create educational materials about AI for kids.

AI Helps Diagnose Multiple Sclerosis

Researchers at Uppsala University have developed an AI model that can determine with 90% certainty which variant of multiple sclerosis (MS) a patient has. The model analyzes data from over 22,000 patients in the Swedish MS Registry and can identify the transition from relapsing-remitting MS to secondary progressive MS earlier than traditional clinical diagnosis. This can help patients receive the right treatment and slow down the progression of the disease. The AI model is based on data already collected during regular healthcare visits, such as neurological tests, MRI scans, and ongoing treatments.

AI Model Detects MS Progression Earlier

A new AI model can detect the transition from relapsing-remitting MS to secondary progressive MS much earlier than traditional clinical diagnosis. The model analyzes routine healthcare information and indicates its confidence level for each assessment. In validation tests, it correctly identified disease progression with about 90% accuracy, often earlier than noted in medical records. Early detection can help patients receive more effective treatments sooner and slow disease progression.

AI at RSA Conference

The RSA Conference will feature talks on the use of AI technologies in threat detection, automated vulnerability management, and the ethics of AI in security. Experts predict that AI will drive industry disruption and that businesses need to move away from traditional security approaches to develop a blueprint for how AI technologies can be deployed. The conference will also host an innovation sandbox, highlighting the work of promising security startups.

USF Launches AI and Cybersecurity College

The University of South Florida is launching the Bellini College of Artificial Intelligence, Cybersecurity and Computing to meet the growing demand for tech talent. The college will serve as a launchpad for the next generation of experts in AI, cybersecurity, and computing. Students will be able to pair their studies with computer science, creating tailored skill sets for in-demand careers. The college aims to fuel Tampa Bay's booming tech industry and contribute to the nation's cybersecurity infrastructure.

AI and Our Digital Legacy

Large language models preserve fragments of human cognition, but the past does not settle in neat layers. Instead, it floats around us, fragmented and adrift. These models collapse time, and there is no inherent sense of when an idea was dominant or obsolete. However, certain structures endure, and bias is stubbornly persistent. The ethical dilemma is how to recover and present these cognitive artifacts without sanitizing history or perpetuating harm.

Apollo.io Appoints New Executives

Apollo.io has appointed Marcio Arnecke as Chief Marketing Officer and Adam Carr as Chief Revenue Officer to accelerate AI-powered go-to-market innovation. The new executives bring expertise in scaling high-growth technology companies and transforming go-to-market strategies. Apollo is transitioning from a high-growth startup to a mature enterprise technology company, generating over $100 million in annual recurring revenue.

CU Professor's AI Grant Canceled

The National Science Foundation canceled a grant for a research project to create social media content that helps young people understand how artificial intelligence works. The grant was awarded to Casey Fiesler, a professor at the University of Colorado, but was abruptly canceled without explanation. The project aimed to create educational materials to help kids understand AI, but the NSF's decision has raised concerns about the impact on STEM education.

AI Model Picks Service Dog Candidates

A team of researchers has found that an AI model is better than humans at assessing puppies for characteristics of good service dogs. The model was trained on data from expert dog trainers and achieved a success rate of 80%. The AI model can help reduce the number of failures in service dog training and improve the selection process. The study was published in the journal Applied Animal Behaviour Science.

US Pushes EU to Delay AI Regulations

The US government is challenging the European Union's upcoming guidelines for advanced artificial intelligence systems, warning that the proposed framework could create undue burdens for developers and stifle innovation. The US Mission to the EU sent a letter to the European Commission, urging a reevaluation of the code of practice currently under development. The EU's Artificial Intelligence Act could result in substantial penalties for noncompliance, and US officials argue that the voluntary framework exceeds the scope of the law.

Conservative Activist Sues Meta over AI Claims

Conservative activist Robby Starbuck is suing Meta over false responses generated by its AI chatbot, including claims that he was a Holocaust denier and participated in the Jan. 6 Capitol riot. Starbuck is seeking over $5 million in damages and alleges that Meta's AI made false claims about him. The lawsuit raises questions about the responsibility of tech companies for AI-generated content and the potential risks of error-prone AI.

Sources

Artificial Intelligence Multiple Sclerosis AI Model Machine Learning Cybersecurity AI Ethics