Microsoft's AI Quake II Demo, Meta's Llama 4 Models, and NxtGen's $300M Funding Quest - Daily AI Brief #149

April 07, 2025 - Daily AI Brief
Hi there, Welcome to today's AI Brief. PwC, Harvey, and Lefebvre have launched a new AI platform aimed at enhancing legal and tax advisory services. Meanwhile, Metro Bank and Ask Silver have introduced a scam detection tool to bolster financial security. In other developments, Microsoft has unveiled an AI demo of Quake II, and Meta has released its latest Llama 4 AI models. Stay tuned for more updates on these and other stories in today's edition.

Top stories

PwC, Harvey, and Lefebvre Launch AI Platform for Legal and Tax Advisory
On April 7, 2025, PwC Tax & Legal, Harvey, and Lefebvre will introduce a new AI platform designed to innovate the legal and tax advisory sectors in Spain. Read more
Metro Bank and Ask Silver Launch Scam Detection Tool
Metro Bank has partnered with AI startup Ask Silver to introduce a scam detection tool to combat fraud in the UK. Customers can upload suspicious communications for analysis and receive feedback via WhatsApp. Read more
PromptArmor Launches AI Vendor Security Testing for Lawyers
California-based startup PromptArmor has introduced a service to assess security risks in generative AI vendors, targeting the legal and health sectors with detailed risk reports. Read more
Meta's AI Model Maverick Faces Benchmark Criticism
On April 6, 2025, Meta introduced its AI model Maverick, which achieved a high ranking on LM Arena. However, discrepancies between the tested and public versions have led to concerns about benchmark reliability and real-world performance. Read more
Microsoft Launches AI Demo of Quake II
Microsoft has introduced a browser-based tech demo of Quake II using its Copilot AI platform. The demo highlights some limitations, including blurry enemies and object permanence issues, and is intended as a research exploration. Read more
Meta Launches Llama 4 AI Models
On April 5, 2025, Meta introduced Llama 4, featuring new AI models Scout, Maverick, and Behemoth. These models are designed to enhance multimodal capabilities and include features to reduce bias and improve response quality. Read more
Google Unveils 'Sec-Gemini v1' for Cybersecurity
Google has launched 'Sec-Gemini v1,' an AI model aimed at enhancing cybersecurity by integrating advanced AI with real-time threat knowledge. The model is available for free to support research and collaboration in the cybersecurity field. Read more

Other news

MOGOPLUS has partnered with Google Cloud to enhance AI lending insights, while Jabil Inc. introduces a 1.6T pluggable transceiver for AI connectivity. pyannoteAI raises $9 million to advance Speaker Intelligence, and A&O Shearman with Harvey launches AI agents for legal workflows. Hon Hai reports a 24% sales increase due to AI demand, contrasting with Samsung's expected 21% profit drop from weak AI chip sales. Knight-Swift partners with Netradyne for AI fleet safety, and Exabeam unveils an AI agent for cybersecurity.

Subscribe to Daily AI Brief

Daily report covering major AI developments and industry news, with both top stories and complete market updates