NerdNews

Apple has announced the iPhone 17e and iPad Air M4, with the iPhone 17e featuring a 6.1-inch Super Retina XDR display, A19 chip, and MagSafe charging. The iPad Air M4 has a 30% faster M4 chip, 12GB RAM, and Wi-Fi 7 support. Other expected announcements include the MacBook A18 Pro, MacBook Pro M5, and Studio Display 2.

OpenAI launched GPT-5.3-Codex-Spark, a production AI model deployed on Cerebras wafer-scale chips, offering improved throughput and low-latency, enabling real-time coding assistance. The model runs at 1,000 tokens per second, 15x faster than earlier versions, and retains its ability to handle long-running processes.

ChatGPT's mobile app uninstalls surged 295% after its deal with the Department of Defense, while competitor Anthropic's Claude app saw a 37% increase in downloads. This suggests a backlash against ChatGPT's partnership with the DoD, with some consumers favoring Anthropic's stance against working with the US defense department.

OpenAI's CEO Sam Altman faced backlash after announcing a Pentagon contract, while Anthropic was blacklisted for refusing similar terms. The incident highlights the challenges AI companies face when working with the US government, particularly in regards to surveillance and automated weaponry. The situation has sparked debate over the role of private companies in national security and the need for clearer guidelines on their collaboration with the government.

Apple's new iPhone 17e features the A19 chip, supports Apple Intelligence, and includes a 48-megapixel camera. It also has MagSafe and Qi2 wireless charging, and is available in black, white, and soft pink. The phone starts at $599 and will be available on March 11.

Huawei is expanding its AI datacenter offerings globally, claiming it can build and deploy them quickly. The company's Intelligent Computing Platform includes its own storage, servers, and GPUs, but its performance lags behind competitors like Intel, AMD, and Nvidia. Despite this, Huawei may find willing buyers in countries that are not prioritized by other suppliers.

Uber's engineering team transformed its data replication platform to move petabytes of data daily across hybrid cloud and on-premise data lakes, addressing scaling challenges caused by rapidly growing workloads. Built on Hadoop's open-source Distcp framework, the platform now handles over one petabyte of daily replication and hundreds of thousands of jobs with improved speed, reliability, and observability.

The CyberStrikeAI tool, an open-source AI security testing platform, has been adopted by hackers to conduct AI-powered attacks. The tool integrates over 100 security tools and allows for automation of attacks, including network scanning, web and application testing, and exploitation frameworks. Researchers warn that this could accelerate automated targeting of vulnerable edge devices.

Deepfakes and injection attacks are evolving, making identity verification more challenging. These threats can bypass traditional security measures, allowing attackers to create fraudulent accounts, take over existing accounts, and gain unauthorized access to sensitive systems. Incode Deepsight offers a solution by validating the entire verification session across three layers: perception, integrity, and behavior.

New AI frameworks like OpenClaw enable tool-using chatbots to be orchestrated in fleets, allowing them to accomplish complex tasks. These fleets can be organized hierarchically and can work 24 hours a day, making them similar to virtual firms. This technology has the potential to revolutionize software engineering and other industries, but also raises security concerns and requires careful management.

Nvidia is investing $4 billion in photonics technology from Lumentum and Coherent to improve data center connectivity and support AI growth. The investment aims to enhance energy efficiency, data transfer speeds, and bandwidth in future AI data centers.

A prayer app with over 5 million downloads was hacked, sending 'surrender' messages to Iranians amid Israeli strikes. The messages, which were sent through push notifications, urged Iranian military personnel to surrender and promised amnesty. The hack is believed to have been planned in advance and is thought to be a nation-state versus nation-state cyber operation.

14.ai, a Y Combinator-backed startup, is replacing legacy customer support teams at many startups it has worked with. The company uses AI-native technology to integrate with support systems, clear ticket backlogs, and monitor tickets across various channels. With $3 million in seed funding, 14.ai aims to increase its headcount and improve its product by experimenting with AI handling most tasks.

Chrome is developing Merkle Tree Certificates (MTCs) to replace traditional digital signatures, reducing authentication data and increasing security against quantum-powered attacks. A three-phase rollout is underway, with the goal of introducing a quantum-resistant root store by 2027.

Asha Sharma, Microsoft's new gaming chief, has stated that she has no tolerance for bad AI in game development. Sharma, who has no prior experience in the gaming industry, faces scrutiny but promises to focus on human-crafted games with deep emotional resonance. Her stance on AI in gaming has sparked debate, with some defending the use of AI tools and others criticizing their implementation.

The Document Foundation has pulled LibreOffice Online out of retirement and is resuming development, sparking controversy over its relationship with Collabora, a company that offers a paid-for cloud-based version of LibreOffice. This move may lead to competition between the two, with LibreOffice Online potentially becoming a free alternative to Collabora's offerings.

Anthropic's Claude AI is experiencing a worldwide outage, with elevated errors reported across all platforms. The issue was flagged on March 2, 2026, and is impacting users broadly, resulting in failed requests, timeouts, or inconsistent responses.

Anthropic's Claude AI service is experiencing a widespread outage, with thousands of users reporting issues accessing the service. The outage affects Claude.ai and Claude Code, but the Claude API is working as intended. The company has identified the issue and is implementing a fix.

Google's Head of Threat Intelligence warns of impending cyber-attacks by Iran against global targets, including the US and Middle East, in response to recent air strikes. Iran's cyber capabilities are expected to target a wide range of organizations, with a focus on hacktivist and ransomware attacks.

Nasdaq is seeking regulatory approval to launch outcome-based options tied to major indexes, allowing traders to place yes-or-no bets on specific outcomes. This move comes as prediction markets like Kalshi and Polymarket gain popularity despite controversy and concerns over insider trading.