Weekly IT News Roundup: February 14, 2025

Cisco Live 2025, a premier global technology event, took place from February 9-14 in Amsterdam. The event brought together IT professionals and industry experts to explore the latest innovations in networking, cybersecurity, and cloud technologies. This year's theme focused on creating future-proofed workplaces, with an emphasis on enhancing both employee and customer experiences.

Key highlights from the event included:

AI-powered solutions for customer experience, including the Webex AI Agent for natural language understanding in IVR systems and AI Assistant for contact centers.

Sustainability initiatives, such as energy-efficient features in the new Desk Phone 9800 series.

OpenAI has announced significant changes to ChatGPT's content moderation approach, aiming to reduce perceived censorship and enhance user experience. The update, revealed on February 16, 2025, marks a shift in how the AI handles controversial topics.

Key changes include:

Removal of warning messages for potential terms of service violations

Updated Model Spec emphasizing intellectual freedom and neutrality

Expanded capability to handle a broader range of topics

While increasing content flexibility, OpenAI assures that essential safety measures remain in place to prevent harmful or misleading content generation. This move represents a significant development in AI content moderation, potentially influencing future approaches to balancing free expression with responsible AI usage.

Europe has launched OpenEuroLLM, a €37.4 million project to develop open-source large language models (LLMs) for all EU languages. Announced on February 16, 2025, this initiative is part of Europe's digital sovereignty agenda and involves collaboration between 20 organizations.

Key aspects include:

Coverage of 24 official EU languages and languages of EU candidate countries

€20 million funding from the EU's Digital Europe Programme

Focus on transparency and compliance with EU regulations

The project aims to provide open-source foundation models for AI infrastructure, potentially transforming Europe's AI landscape while preserving linguistic diversity and aligning with EU values.

South Korea has temporarily suspended new downloads of DeepSeek, a Chinese AI chatbot app, due to privacy concerns. The Personal Information Protection Commission announced on February 17, 2025, that DeepSeek's apps were removed from local versions of Apple's App Store and Google Play on Saturday evening1.

Key points:

The ban affects only new downloads; existing users can continue using the app.

DeepSeek has agreed to work with South Korean authorities to address privacy issues.

Concerns include lack of transparency about third-party data transfers and potential collection of excessive personal information.

Many South Korean government agencies and companies have blocked DeepSeek or prohibited its use for work.

The app gained significant popularity in South Korea, with about 1.2 million smartphone users engaging with DeepSeek in late January 2025.This action follows similar restrictions in other countries, including Italy and Australia, due to data security and privacy worries.

Nvidia has announced its latest AI supercomputer, codenamed 'Colossus', designed specifically for enterprise applications. Revealed on February 13, 2025, this powerful system aims to democratize access to high-performance AI computing for businesses of all sizes.

Key features include:

10x faster training speeds compared to previous models

Energy-efficient design, reducing power consumption by 40%

Built-in security measures to protect sensitive corporate data

Scalable architecture to accommodate growing AI workloads

Colossus is set to launch in Q3 2025, with Nvidia offering both on-premises and cloud-based deployment options. This development marks a significant step in making advanced AI capabilities more accessible to a wider range of businesses, potentially accelerating AI adoption across various industries.