Nothing CEO Carl Pei has announced a new screenless AI-native device set to launch in December. The product will have no display and could not have been built as an app, according to Pei. Details remain scarce, but reports suggest it may be an ambient recorder or note-taker designed to capture fleeting ideas and conversations before they are forgotten. A teaser image showed a textured, nearly featureless corner of the device.
India's IT ministry has released a draft framework requiring AI models to be verified before deployment. The proposal includes model vetting and safeguards across the training process. Industry experts are concerned it could function like a license and tighten over time. Co-founder Sharad Sharma said the goal is to build trust and accountability in AI development.
Legal AI has grown rapidly since late 2022, with major moves from OpenAI, Google, Anthropic and Microsoft. Thomson Reuters acquired Casetext and launched its own LLM called Thomson. Harvey reached over $15 billion in valuation, while Legora grew past $1.4 billion in total investment. AI-native law firms have surged from a handful to several dozen in 2026.
Key Takeaways
- Nothing is launching a screenless AI-native device in December, led by CEO Carl Pei
- India proposes mandatory AI model vetting before deployment, raising concerns about overregulation
- Harvey reached over $15 billion in valuation in the legal AI space
- Thomson Reuters launched its own LLM called Thomson after acquiring Casetext
- Legora surpassed $1.4 billion in total investment
- AI-native law firms have grown from a few to several dozen in 2026
- Kent newsletter network uses Claude, Airtable and Stripe while keeping human editors
- Augusta University develops SE Tutor AI with $2 million NSF grant for software engineering students
- AI pentesting tool ARTEMIS placed second against ten professional testers on 8,000 hosts
- Kablamo appoints Nat Abood, ex-Xero, as Head of Product and AI Strategy
Nothing reveals screenless AI device coming in December
Nothing CEO Carl Pei announced a new screenless AI-native product launching in December. He said the device could not have been an app and will have no display. Reports suggest it may be an ambient recorder for capturing everyday ideas and conversations. Pei explained that important moments often slip away because people do not record them in time.
Nothing teases mysterious AI hardware launch next month
Nothing is set to release a new AI-native device in December with almost no details shared. CEO Carl Pei posted on X that the product has no screen and could not have been an app. It may be a note-taker or recorder designed to help capture fleeting ideas. Pei also shared a teaser image showing a textured, nearly featureless corner of the device.
India proposes AI model vetting before deployment
India's IT ministry released a draft framework requiring AI models to be verified before use. The plan includes model vetting and safeguards across the training process. Industry experts worry it could act like a license and become stricter over time. Co-founder Sharad Sharma said the goal is to build trust and accountability in AI development.
A brief history of generative AI in law
Legal AI has grown rapidly since late 2022, with major moves from OpenAI, Google, Anthropic and Microsoft. Thomson Reuters bought Casetext and later launched its own LLM called Thomson. Harvey reached over $15 billion in valuation, and Legora grew past $1.4 billion in total investment. AI-native law firms have surged from a handful to several dozen in 2026.
Kent newsletter network scales with AI but keeps human editing
Don't Miss Media Group grew its Kent newsletters to over 6,000 subscribers across Margate, Ramsgate and Broadstairs. The company uses AI tools like Claude, Airtable and Stripe for admin, sales and formatting tasks. Human editors still handle reporting, curation and all editorial decisions. Founder Georgina Wilson-Powell said AI helps efficiency but cannot replace local voice and judgement.
Augusta University develops AI tutor for software engineering
Augusta University researchers are creating the Software Engineering (SE) Tutor, an AI learning tool funded by a $2 million grant from the U.S. National Science Foundation. The tool uses generative AI to turn technical feedback into guided questions that help students understand and fix coding mistakes. It aims to reach over 2,000 students and 50 faculty members across three universities. The project plans to release SE Tutor as open-source software with free training materials for educators.
AI pentesting changes security testing for DevSecOps teams
AI pentesting is transforming security testing by automating repetitive tasks like probing endpoints and retesting after releases. It can investigate attack paths that traditional scanners miss and identify dangerous misconfigurations that do not have CVE entries. AI pentesting enables continuous security testing within the software delivery lifecycle rather than only at major release points. A December 2025 study showed an AI agent called ARTEMIS placed second against ten professional testers on a network of about 8,000 hosts. Human testers remain important for complex judgment-based work while AI handles high-volume tasks.
AI prescribing could reshape pharmacist roles in health care
States are beginning to allow AI systems to write routine prescriptions, which could reduce tedious work for pharmacists. In Utah's pilot program, Nolla Health's AI application scores acne severity from facial images and initiates prescription orders with physician oversight. A national study found that 88% of chain pharmacists have more work activities than when hired and 85% say workload metrics compromise patient safety. Experts say AI will enhance rather than replace pharmacists because complex patient counseling still requires human judgment. Regulatory sandbox programs include Nolla Health, August AI, and Expect Fitness.
Anti-AI products top Japan's 2027 hit predictions
Japanese magazine Nikkei Trendy ranked the top products expected to become hits in 2027, with anti-AI as the central theme. The top prediction is a human certificate, a digital system that proves a person is real and prevents online impersonation. Edible cactus also ranked highly for its nutritional value and cooking versatility. The magazine's 2026 hit list included affordable items like decorative marker charms and Bonbon Drop stickers. These rankings show Japanese consumers are seeking authenticity and small affordable pleasures amid rising prices.
Kablamo appoints Nat Abood as head of product and AI strategy
Kablamo named Nat Abood as Head of Product and AI Strategy. Abood previously spent five years at Xero as Principal Product Manager leading agentic AI tools. He will oversee AI and agentic solutions for Kablamo's clients and the company itself, based in Melbourne. Before Xero, Abood worked at REA Group, Origin Energy, and led Origin's first chatbots and voice assistants with Google. Co-CEO Allan Waddell said the appointment reflects Kablamo's focus on helping businesses apply AI to operations with clear governance and measurable results.
Sources
- Nothing teases screenless AI product for December, says it ‘couldn't have been an app’
- Nothing teases a mysterious ‘AI-native product’ launch for next month
- MeitY’s AI training framework proposes model vetting, raises questions over approval norms
- A Brief History of Generative Legal AI
- Kent newsletter network uses AI to scale without handing over editorial judgement
- Advancing software engineering education with artificial intelligence
- AI pentesting is reshaping security testing for DevSecOps
- How AI Prescribing Could Transform the Pharmacist’s Role in Health Care
- Anti-AI Tops Hit Product Predictions
- Kablamo names Nat Abood head of product & AI strategy
Comments
Please log in to post a comment.