From record AI funding and multimodal model launches to malware campaigns, robotaxi failures, and fresh waves of layoffs, this week’s tech news read like a speedrun through the industry’s biggest ambitions and messiest vulnerabilities. Together, these stories capture a sector pushing hard into the future while still tripping over security gaps, platform risks, and the human cost of constant reinvention.
Top news
AI innovation accelerates across the globe
Alibaba unveiled its Qwen3.5 Omni model, a multimodal AI capable of processing text, images, audio, and video in over 100 languages. The release introduces a new architecture for smoother speech synthesis but signals a move away from open-source distribution.
Microsoft launched its MAI Suite, featuring transcription, voice, and image models, while Google released the open-source Gemma 4 family under the Apache 2.0 license. Alibaba also expanded its lineup with Wan2.7-Image and Qwen 3.6-Plus.
Anthropic faced a major leak revealing “Claude Mythos,” an advanced AI model surpassing Opus 4.6 in reasoning and cybersecurity. The company confirmed early testing and attributed the breach to a human error involving an unsecured CMS.
Anthropic also suffered a separate leak when its Claude Code CLI source was accidentally published via npm, exposing over 500,000 lines of TypeScript. The company issued thousands of DMCA takedowns and confirmed no customer data was compromised.
OpenAI raised a record-breaking $122 billion at an $852 billion valuation, with backing from Amazon, Nvidia, and SoftBank. The company plans to merge its tools into a single AI superapp and is reportedly preparing for an IPO later this year.
Consumer and developer tools evolve
Google now allows users to change the username portion of their Gmail addresses without losing data. The feature, rolling out in the US, includes limits on frequency and total number of changes.
Google introduces stronger privacy controls in Android 17 Beta 3, including single-session location sharing and post-quantum APK signing. Developers are encouraged to update apps for compatibility with the new standards.
Google also added one-click chat migration tools to Gemini, allowing users to import conversations from ChatGPT and Claude. The update supports up to 5 GB of data and rebrands the “Chats” pane to “Memory,” though enterprise and EU users are excluded.
Apple is said to be opening up iOS 27’s Siri to third-party AI chatbots like Google Gemini and Anthropic Claude. A new “Extensions” menu will let users assign specific tasks to different AIs, marking a major shift from Siri’s exclusive ChatGPT integration.
Robotics and automation take center stage
AgiBot surpassed 10,000 humanoid robot deliveries, doubling production in just 90 days. The Shanghai-based company now leads the global commercial humanoid market, outpacing Tesla and Agility Robotics.
Baidu faced a major setback when a system failure halted over 100 Apollo Go robotaxis in Wuhan, stranding passengers and causing traffic jams. The incident reignited global debates about autonomous vehicle safety.
Figure AI founder Brett Adcock launched Hark, a new company developing personal AI devices for home use. Backed by $100 million, Hark is building its own multimodal AI model and plans to release hardware later this year.
Insider intel
Jamie Dimon, CEO of JPMorgan Chase, predicted that AI-driven productivity gains could enable future generations to work just 3.5 days per week. He warned of potential job disruption and urged coordinated reskilling efforts, echoing similar sentiments from other tech leaders.
Security alerts
Corporate and consumer breaches
Hasbro reported a cyberattack that forced key systems offline, delaying shipments and licensing deals. The company is working with external responders and advising partners to monitor communications closely.
NoVoice, a rootkit malware campaign, infected up to 2.3 million Android devices through over 50 apps on Google Play. Google has since removed the affected apps and urged users to update their devices.
WhatsApp users were targeted by a new malware campaign using Visual Basic Script attachments disguised as Windows utilities. The attack installs a stealth backdoor, prompting warnings to block unknown files and maintain updated antivirus protection.
Apple and Google security updates
Apple released iOS 18.7.7 to patch the DarkSword exploit affecting older iPhones. The update fixes a Safari hijack vulnerability that could steal crypto and personal data.
Google rolled out AI-based ransomware detection on Google Drive to all Workspace tiers, automatically halting syncs and alerting users when suspicious encryption is detected.
Apple now includes a Terminal safeguard in macOS 26.4 that detects and blocks suspicious pasted commands, protecting users from scams like ClickFix. The feature warns users before execution but can be bypassed manually.
High-profile breaches and fixes
OpenAI patched a ChatGPT vulnerability that could leak data via DNS lookups, potentially exposing chat histories and files. A separate Codex flaw exposing GitHub tokens was also resolved earlier this year.
Iranian hackers breached the personal Gmail account of FBI Director Kash Patel, leaking over 300 emails. The FBI confirmed that no classified data was exposed, and the US is offering a $10 million reward for information on the attackers.
Industry shakeups
IPO ambitions and financial moves
SpaceX confidentially filed for an IPO that could exceed a $2 trillion valuation, potentially the largest in history. The company plans a dual-class share structure and aims to attract retail investors while leveraging Starlink and xAI growth.
Corporate restructuring and workforce shifts
Oracle announced layoffs affecting up to 30,000 employees as part of a $2.1 billion restructuring to fund a $50 billion AI infrastructure expansion. Despite concerns about long-term risks, investors responded positively.
Apple is offering stock bonuses worth up to $400,000 to retain iPhone design staff amid aggressive recruiting by OpenAI and startups like Hark. The move underscores Apple’s struggle to keep top talent as the market pivots toward AI wearables.
Service disruptions and regulatory changes
DeepSeek, China’s top ChatGPT rival, experienced a seven-hour outage affecting 355 million users. Analysts cited potential code or capacity issues as the company prepares for its next major model release.
Apple also rolled out iOS 26.4 in the UK, enforcing age verification for adult-rated apps and unfiltered content. The update aligns with the UK’s Online Safety Act and uses ID scans or credit card checks for verification.
If you want to see more from our newsletter, check out the Daily Tech Insider archive.
Read the full article here
