This week brought a mix of developments across the technology sector, ranging from artificial intelligence tools for government and everyday users to rising cybersecurity threats and new advancements in AI hardware. These updates reflect how large tech firms are expanding their services, competing for government and commercial clients, and responding to global digital threats.
Google has introduced a new feature for its Gemini AI that enables video upload analysis directly through the mobile app. Users can now upload video files and receive frame-by-frame interpretation based on their queries. The tool works without requiring any subscription, needing only an updated version of the app, available on iOS and Android. With this feature, Google aims to improve how users interact with AI by making it useful for tasks like identifying objects or analysing movement in recorded footage. This development puts Google in closer competition with OpenAI, which already offers multimodal AI features. However, questions around how user-uploaded videos are processed and stored are expected to spark debates over privacy and data handling.
In a parallel AI push, Microsoft has built a customised version of its Copilot AI for the United States Department of Defence. This version is scheduled for rollout in the summer of 2025 and will be used by over one million personnel. It is designed to handle secure, high-stakes environments such as battlefield coordination and logistics planning. The move places Microsoft at the centre of growing military interest in AI technologies. Alongside this, OpenAI’s $200 million agreement with the Pentagon suggests that AI will play a growing role in government systems. Critics are raising questions about ethical boundaries and the risks of automation in military operations, particularly when dealing with sensitive or classified data.
Hardware developments also made headlines, with Dell announcing its new Pro Max desktop PC series tailored for AI research and development. Powered by Nvidia’s Grace Blackwell architecture, the devices promise to deliver high processing power with improved energy efficiency. Set to launch in mid-2025, the series targets enterprises working on large-scale AI models. While Dell’s move strengthens its presence in enterprise computing, the high cost of these systems could limit their availability to smaller firms and independent developers.
The tech world was also reminded of its vulnerability to cyber threats. The Washington Post reported a breach in which hackers targeted journalists’ email accounts. Around the same time, Canadian airline WestJet confirmed disruptions to both its mobile app and internal systems. The incidents, reported between June 14 and 15, raised concerns about weak cybersecurity defences across critical sectors. Experts suspect that state-backed hackers may be involved, though no official blame has been assigned. Both companies have responded by upgrading their internal security protocols.
At the SAFE Forum in San Jose, Samsung Foundry and Groq shared their plan to mass-produce a new language processing unit (LPU), claiming it to be the fastest AI chip under development. Built on Samsung’s 4nm process, the chip is aimed at industries such as robotics, autonomous transport, and high-speed communication systems. Its expected release in late 2025 could challenge Nvidia’s leading position in AI chips. However, supply chain issues and production costs may become challenges as the chip nears its market debut.