Latest news with #USDepartmentofDefence
India Today
a day ago
- Business
- India Today
Microsoft stops using Chinese tech support for this customer over national security concerns
Microsoft has announced a major policy shift, cutting off technical support from its China-based engineering teams for US defence clients using its cloud services. The move comes just days after an explosive ProPublica investigation revealed how the US Department of Defence (DoD) relied on Microsoft software engineers located in China, a situation that sparked concerns over national security and cyber a statement posted to X on Friday, Frank Shaw, Microsoft's Chief Communications Officer, confirmed the change: 'In response to concerns raised earlier this week about US-supervised foreign engineers, Microsoft has made changes to our support for US Government customers to assure that no China-based engineering teams are providing technical assistance for DoD Government cloud and related services.' The policy revision directly impacts Microsoft Azure, the company's cloud services division, which now accounts for over 25 per cent of Microsoft's global revenue, larger than Google Cloud but still behind Amazon Web Services. The company's most recent earnings report noted that more than half of its $70 billion in Q1 revenue was generated from US-based clients, with a substantial chunk tied to government isn't the first time Microsoft's cloud involvement with the US military has drawn scrutiny. Back in 2019, the company secured a $10 billion cloud contract with the Pentagon, only for it to be cancelled in 2021 after legal disputes. By 2022, Microsoft was again one of the chosen vendors in a $9 billion multi-supplier defence cloud deal alongside Amazon, Google, and ProPublica report, which prompted the recent policy overhaul, revealed that Chinese Azure engineers were managed remotely by US-based supervisors known as 'digital escorts', individuals who reportedly had less technical knowledge than the staff they were overseeing. According to the report, this system could leave critical US defence infrastructure vulnerable to foreign Defence Secretary Pete Hegseth didn't mince words in response. 'This is obviously unacceptable, especially in today's digital threat environment,' he said in a video posted on X. Hegseth criticised the structure as 'a legacy system created over a decade ago, during the Obama administration,' and confirmed the Department of Defence would review its infrastructure for similar on DOD's cloud services Secretary of Defense Pete Hegseth (@SecDef) July 18, 2025While Microsoft initially defended its operations, claiming that its teams followed US regulations, it has since promised to undergo further scrutiny and cooperate with national security partners. 'We remain committed to providing the most secure services possible to the US government, including working with our national security partners to evaluate and adjust our security protocols as needed,' Shaw incident highlights the growing tension between global tech operations and national security expectations, a balancing act that's becoming increasingly difficult in today's high-stakes cyber environment.- Ends
Sydney Morning Herald
4 days ago
- Sydney Morning Herald
‘MechaHitler': Why Elon Musk's chatbot is at the centre of an Australian legal dispute
Australia's online safety watchdog is back in court this week, battling Elon Musk's X over issues of AI, free speech and who is ultimately responsible for detecting and removing violent online content. What is Grok AI, and why has it been controversial? Elon Musk's AI chatbot, dubbed Grok, is embedded in X (formerly Twitter) and has made headlines for numerous controversies – as well as for winning a $US200 million ($300 million) contract with the Pentagon. It's arguably more capable at present than ChatGPT and Gemini, but has proven much more unpredictable too. Last week, Grok declared itself a super-Nazi, referring to itself as 'MechaHitler', and made racist, sexist and antisemitic posts that its parent company, xAI, later deleted. Musk has said he wants Grok to 'not shy away from making claims which are politically incorrect', and this seems to be the result. Then, just days later, xAI launched a girlfriend chatbot that is available to 12-year-olds, despite being programmed to engage in sexual conversation. Loading The sexualised AI chatbot, named Ani, communicates with emojis and flirtatious messages and can appear dressed in lingerie. It's designed to act as if it is 'crazy in love' and 'extremely jealous', according to programming instructions posted on social media. Despite all this, xAI announced it had won a contract worth up to $US200 million to develop artificial intelligence tools for the US Department of Defence. If the chaos proves anything, it's that AI chatbots such as Grok are moving incredibly quickly, and regulators and governments are racing to catch up.
Focus Malaysia
4 days ago
- Business
- Focus Malaysia
Eyes in the sky: Strengthening Malaysia's defence with AI and drones
AS THE Indo-Pacific region experiences growing instability driven by great power competition, especially in the South China Sea, Malaysia must adapt its defence strategy to respond to emerging threats. Rather than pursuing costly conventional weapons, Malaysia has a unique opportunity to enhance its military readiness through the integration of artificial intelligence (AI) and drone technology. This dual strategy provides a cost-effective, agile, and non-escalatory means of securing national interests while preserving its diplomatic neutrality and regional credibility. Malaysia's defence focus has already begun shifting toward unmanned aerial vehicles (UAVs). In 2025, the Royal Malaysian Air Force acquired three Turkish-made Anka-S drones i.e., medium-altitude, long-endurance (MALE) UAVs equipped with synthetic aperture radar, electro-optical/infrared sensors, and secure satellite communications. Stationed in Labuan, these drones are unarmed and used primarily for surveillance over Malaysia's Exclusive Economic Zone. This approach allows Malaysia to assert maritime sovereignty, especially in disputed waters, without provoking regional tensions. Malaysia is also working to localise drone development by collaborating with nations such as South Korea and China, while developing domestic UAV production capabilities through partnerships with universities and technical institutions. The goal is to manufacture thousands of drones annually, integrating them into both civilian and military uses. However, drones alone are not enough. The real transformative potential lies in integrating AI into military operations. AI-driven systems can process vast streams of sensor data in real time, enabling rapid detection, classification, and response to potential threats. These include identifying hostile vessels or aircraft, tracking illegal fishing, or detecting suspicious maritime activity. AI can also coordinate multiple drones to operate in autonomous swarms sharing tasks like surveillance, signal jamming, or area denial. This swarm capability provides a force multiplier, allowing Malaysia to cover wide areas of ocean with minimal human input. Further, AI can be used in decision-support systems for military commanders. Inspired by the US Department of Defence's 'Thunderforge' initiative where Scale AI is a key partner, Malaysia could explore similar AI tools to assist in generating operational plans, simulating battlefield scenarios, and prioritising threats, all while ensuring that human decision-makers remain fully in control. This kind of smart command-and-control infrastructure ensures better preparedness and quicker responses, especially in contested or fast-changing environments like the South China Sea. Neighbouring countries offer useful models. Taiwan has developed loitering drones with AI targeting, India's 'Akashteer' programme uses AI to coordinate counter-drone warfare across hundreds of kilometres, and AUKUS nations have tested multi-drone systems that adapt in real-time to electronic warfare. While Malaysia has not yet embraced armed drones, these developments show the technological direction of regional militaries. For Malaysia, a measured approach prioritising surveillance and deterrence while staying within international law is both viable and strategically wise. To fully realise these capabilities, Malaysia must develop a strong national AI-drone ecosystem. This means investing in homegrown research, engineering, and manufacturing. Drone-specific AI frameworks should be developed in collaboration with defence contractors, start-ups, and universities. Ethical frameworks and governance structures must be established to ensure that AI systems remain under meaningful human control, comply with humanitarian law, and avoid misuse. ASEAN's ongoing efforts to shape regional AI norms can guide Malaysia's approach, balancing innovation with responsibility. The future of defence is not just about firepower, it's about information, intelligence, and interoperability. Malaysia must ensure that its drones and AI systems are not siloed, but integrated across naval, air, and ground forces. For example, drone swarms could relay real-time video to navy vessels, while AI systems track and flag suspicious movement along maritime borders. Edge computing devices onboard drones can process data locally, minimising delays caused by remote command centres. This kind of seamless, multi-domain integration is crucial for fast, adaptive operations in a crisis. Importantly, Malaysia must not lose sight of its strategic identity as a neutral and peace-oriented nation. Drones and AI should not be tools of provocation but instruments of vigilance and deterrence. By maintaining unarmed drone deployments, emphasising surveillance over strike capabilities, and using AI for defensive planning, Malaysia can strengthen its position without becoming entangled in the escalating militarisation of the region. This approach allows Malaysia to assert its rights under international law, monitor its waters against encroachment or criminal activity, and coordinate better with regional partners on maritime safety. In the long run, Malaysia can take the lead in shaping an ASEAN-wide doctrine for AI and drone cooperation. This could include shared surveillance protocols, disaster response coordination, and collective counter-UAV strategies. By promoting interoperability, transparency, and civilian-military integration, Malaysia would contribute to regional stability without becoming dependent on any external bloc. As global militaries enter the age of autonomous systems and algorithmic warfare, Malaysia must not be left behind. Through careful investment in AI, drone technology, and regional cooperation, Malaysia can build a modern defence force that is smart, agile, and ethically grounded. The goal is not domination but assurance guarding sovereignty with intelligence, and preparing for conflict without seeking it. In the contested waters and skies of the Indo-Pacific, Malaysia's best defence may lie not in missiles, but in the sharp eyes and fast minds of AI-guided drones. ‒ July 17, 2025 R. Paneir Selvam is the principal consultant of Arunachala Research & Consultancy Sdn Bhd, a think tank specialising in strategic national and geopolitical matters. The views expressed are solely of the author and do not necessarily reflect those of Focus Malaysia. Main image: Defence News
Sydney Morning Herald
5 days ago
- Sydney Morning Herald
Why are X and the eSafety Commissioner back in court?
Australia's online safety watchdog is back in court this week, battling Elon Musk's X over issues of AI, free speech and who is ultimately responsible for detecting and removing violent online content. What is Grok AI, and why has it been controversial? Elon Musk's AI chatbot, dubbed Grok, is embedded in X (formerly Twitter) and has made headlines for numerous controversies – as well as for winning a $US200 million ($300 million) contract with the Pentagon. It's arguably more capable at present than ChatGPT and Gemini, but has proven much more unpredictable too. Last week, Grok declared itself a super-Nazi, referring to itself as 'MechaHitler', and made racist, sexist and antisemitic posts that its parent company, xAI, later deleted. Musk has said he wants Grok to 'not shy away from making claims which are politically incorrect', and this seems to be the result. Then, just days later, xAI launched a girlfriend chatbot that is available to 12-year-olds, despite being programmed to engage in sexual conversation. Loading The sexualised AI chatbot, named Ani, communicates with emojis and flirtatious messages and can appear dressed in lingerie. It's designed to act as if it is 'crazy in love' and 'extremely jealous', according to programming instructions posted on social media. Despite all this, xAI announced it had won a contract worth up to $US200 million to develop artificial intelligence tools for the US Department of Defence. If the chaos proves anything, it's that AI chatbots such as Grok are moving incredibly quickly, and regulators and governments are racing to catch up.
The Age
5 days ago
- The Age
Why are X and the eSafety Commissioner back in court?
Australia's online safety watchdog is back in court this week, battling Elon Musk's X over issues of AI, free speech and who is ultimately responsible for detecting and removing violent online content. What is Grok AI, and why has it been controversial? Elon Musk's AI chatbot, dubbed Grok, is embedded in X (formerly Twitter) and has made headlines for numerous controversies – as well as for winning a $US200 million ($300 million) contract with the Pentagon. It's arguably more capable at present than ChatGPT and Gemini, but has proven much more unpredictable too. Last week, Grok declared itself a super-Nazi, referring to itself as 'MechaHitler', and made racist, sexist and antisemitic posts that its parent company, xAI, later deleted. Musk has said he wants Grok to 'not shy away from making claims which are politically incorrect', and this seems to be the result. Then, just days later, xAI launched a girlfriend chatbot that is available to 12-year-olds, despite being programmed to engage in sexual conversation. Loading The sexualised AI chatbot, named Ani, communicates with emojis and flirtatious messages and can appear dressed in lingerie. It's designed to act as if it is 'crazy in love' and 'extremely jealous', according to programming instructions posted on social media. Despite all this, xAI announced it had won a contract worth up to $US200 million to develop artificial intelligence tools for the US Department of Defence. If the chaos proves anything, it's that AI chatbots such as Grok are moving incredibly quickly, and regulators and governments are racing to catch up.
