Microsoft has introduced new data and artificial intelligence products to help healthcare organizations access and analyze vast amounts of healthcare data generated daily. These products aim to address the challenge of unifying data from various sources like electronic health records, medical devices, lab systems, and images, making it easier for healthcare professionals to access and standardize this information for decision-making. Microsoft is trialing these tools with healthcare providers and hopes they will help improve patient care and optimize healthcare systems. Additionally, Microsoft unveiled Azure AI Health Bot, Text Analytics for health, and Azure AI Health Insights models designed to aid healthcare providers and researchers in making informed decisions and simplifying patient data. These solutions are available in a preview capacity. Read More
ServiceNow, a major software company, has introduced the "Now Assist" family of solutions with generative AI capabilities. This enhancement spans IT Service Management (ITSM), Customer Service Management (CSM), Human Resources Service Delivery (HRSD), and application development. For ITSM, Now Assist provides comprehensive incident summaries and real-time interactions with virtual agents, improving issue resolution. In CSM, it streamlines customer service by generating summaries for cases and chat interactions, enhancing agent productivity. For HRSD, Now Assist streamlines HR operations, reducing manual tasks and providing quick answers for employees. Now Assist for Creators supports rapid application development on the Now Platform. It includes a text-to-code feature, simplifying coding processes. This AI enhancement solidifies ServiceNow's commitment to innovation and strengthens its position in digital workflow automation. Read More
Swiss startup Lakerais launching a service designed to protect enterprises from various security weaknesses inherent in large language models (LLMs). While capable of generating human-language text from prompts, malicious users can manipulate these LLMs using "prompt injection" techniques to achieve unauthorized access or bypass security measures. Lakera, which raised $10 million in funding, developed insights from an interactive game called Gandalf, in which users attempt to trick an LLM into revealing a secret password. Lakera Guard, the startup's flagship product, offers protection against prompt injections and data leakage. It can also protect against confidential data leaking into the public domain and moderate content to ensure LLMs do not produce unsuitable content. The company also protects against LLM-enabled misinformation and factual inaccuracies, addressing "hallucinations" when LLM output contradicts system instructions or delivers factually incorrect information. Lakera's launch aligns with the EU AI Act, focusing on safeguarding generative AI models. The company has served in advisory roles to the Act, assisting in establishing technical foundations. Read More