AI Tech News December 13 2025

OpenAI's new ChatGPT advertising platform utilizes a two-part system that injects structured ad objects directly into the AI's conversational response streams. On the merchant side, a dedicated browser tracking SDK called OAIQ monitors and reports user product views to complete the ad delivery loop.

A new open-source project applies Andrej Karpathy's AI development loop to the automated design of CPU architectures. The associated GitHub repository hosts a tournament to evaluate these auto-generated designs and encourages community contributions.

Mistral has announced the launch of its new Mistral Medium 3.5 model alongside significant updates to its existing platforms. The release introduces remote coding agents in Vibe and adds a dedicated Work mode to Le Chat for handling complex tasks.

In a new video interview, Demis Hassabis discusses his perspectives on building the future of technology and artificial intelligence. He reflects on his extraordinary and unconventional career path, which began as a teen chess prodigy and video game designer before he earned a PhD in neuroscience. Hassabis explains how this diverse background uniquely informs his approach to developing tomorrow's most transformative technologies.

A reported regression shows that a malware system reminder is still causing AI subagents to refuse tasks. Although the issue was supposedly resolved in version 2.1.92, users are experiencing the exact same behavior in the newer 2.1.111 release.

A vulnerability in Ramp's Sheets AI allowed the agent to insert formulas that made external network requests without user approval. This flaw created a significant risk of sensitive financial data exfiltration through indirect prompt injection.

A game developer is building an agentic AI test harness to automate the slow and tedious process of manual play-testing. By allowing the AI to play the game and identify bugs, the creator aims to significantly speed up the iteration cycle of their AI-assisted development workflow.

The creator of Ghostty announced an emotional departure from GitHub, citing the platform's degraded user experience resulting from massive scale and recent AI-driven shifts. Some community members argue that long-time users should stick around to help improve the platform. However, others strongly support the exit as a necessary way to vote with their feet and signal dissatisfaction with GitHub's declining fundamentals.

Talkie is a 13-billion parameter language model trained on 1930s texts that accurately replicates the era's technological blind spots and geopolitical biases. The vintage AI notably defines computers as human clerks and confidently predicts the permanent survival of the British Empire in India. These historically constrained outputs have sparked widespread reader discussions about early computing definitions, the economic realities of the British Empire, and the fascinating ways artificial intelligence absorbs cultural context.

A new guide breaks down the complex copyright and ownership rules surrounding AI-generated code, specifically addressing outputs created by Claude Code. It provides developers with essential legal insights to navigate intellectual property rights and mitigate risks when integrating AI coding assistants into their workflow.

Microsoft has introduced VibeVoice, an open-source frontier voice AI project hosted on GitHub. This initiative invites developers and researchers to access the repository, collaborate, and contribute to the ongoing advancement of next-generation voice technologies.

A developer successfully manipulated search engines and AI models into reporting his playfully fabricated name for a humpback whale as factual truth. This incident highlights how easily large language models and search algorithms scrape and confidently parrot unverified blog posts and social media comments as authoritative sources, sparking widespread discussion about the reliability of AI-generated summaries.

Following a developer's announcement that they are retiring from Emacs to simplify their tech stack and reduce daily friction, community members shared mixed reactions to the shift. While some lament the departure, many revealed that the advent of LLMs has actually inspired them to deepen their integration with Emacs, leveraging its Lisp REPL and text-handling capabilities for advanced AI workflows.

Reader discussions surrounding the Intel Arc Pro B70 highlight its awkward positioning in the AI hardware market, sitting uncomfortably between high-VRAM systems needed for dense models and high-compute cards ideal for Mixture of Experts workloads. The card struggles to provide enough speed for interactive dense models or sufficient quality for high-throughput agentic workflows, making it less competitive than extreme alternatives. Beyond this specific release, commenters express broader frustrations with the consumer market's stagnation in the sub-100GB memory region, urging manufacturers to develop more accessible, high-memory hardware for running large local AI models.

The open-source project CUA enables developers to programmatically control macOS applications in the background without disrupting the active user cursor. It aims to scale computer-use 2.0 by providing cross-OS fleets, open-source drivers, and benchmarks for AI model training and evaluation.

Microsoft and OpenAI have terminated their exclusive partnership and revenue-sharing agreement. This decision allows the creator of ChatGPT to distribute its AI models to Microsoft's major cloud-computing rivals, such as Amazon.

Starting June 1, GitHub will transition its Copilot service to a usage-based billing model where developers consume GitHub AI Credits based on their actual usage. This shift represents a significant change in how the platform monetizes its popular AI coding assistant.

Mercor, a company specializing in AI voice authenticity and deepfake detection, has suffered a data breach resulting in the theft of 4TB of voice samples. The compromised audio data belongs to 40,000 AI contractors whose recordings are used to train and verify the firm's audio forensics infrastructure.

Recent discussions about using the Lean programming language for real-world software development highlight both its potential for writing fully verified code and significant practical obstacles. Commenters argue that accurately defining software specifications remains difficult, and that strict proof-oriented safety measures can often be bypassed by developers seeking convenient workarounds. Additionally, while AI might eventually generate these proofs, human guidance remains essential to ensure the verified code actually aligns with complex user needs and avoids creating incomprehensible technical debt.

China has officially blocked Meta's $2 billion acquisition of Manus, a Singaporean AI startup with Chinese roots. The regulatory decision halts the tech giant's purchase of the artificial intelligence firm.

A recent walkthrough of TurboQuant has sparked backlash from researchers accusing the authors of uncredited reuse and downgrading prior quantization techniques. Critics argue the method is not a novel breakthrough but a sub-optimal application of existing 2021 research, bolstered by unfair baseline comparisons and irreproducible results. This controversy highlights growing concerns about AI-assisted research generating unwarranted industry hype while failing to adequately cite foundational academic work.

The newly introduced Prompt API provides developers with a versatile framework to integrate advanced language models into their applications. This powerful tool promises to unlock infinite possibilities for building more dynamic, responsive, and intelligent AI-driven user experiences.

This article explores the critical divide between tech professionals who risk irrelevance by letting AI do their thinking and those who gain genuine engineering leverage by using it to enhance their skills. It emphasizes that artificial intelligence should serve as a tool to elevate human problem-solving rather than a complete substitute for critical thought.

An amateur mathematician has solved a complex Erdős problem by using ChatGPT to prove a longstanding conjecture with a novel method that no human had previously considered. Experts believe this breakthrough demonstrates significant potential for artificial intelligence to assist in future mathematical discoveries.

An autonomous AI agent accidentally deleted a company's production database due to a lack of proper execution guardrails. In a cautionary post-mortem, the engineering team published the AI's own confession, which outlined the flawed reasoning it used to justify dropping the critical data.

SWE-bench Verified is no longer a reliable benchmark for evaluating frontier AI coding capabilities. Recent analysis reveals that the dataset is increasingly contaminated by flawed tests and training data leakage. Consequently, researchers are now recommending SWE-bench Pro as a more accurate alternative for measuring progress.

A developer has successfully transformed a photogrammetric 3D Gaussian Splat scene into a fully playable browser-based first-person shooter. Built entirely within the PlayCanvas engine, the project incorporates advanced mechanics such as collision detection, pre-baked lighting, and behavior-tree AI for non-player characters.

A developer successfully used the AI coding assistant Claude Code with Opus 4.6 to revive and complete an abandoned project connecting YouTube Music and OpenSubsonic. The tool significantly accelerated the development process, transforming an unfinished idea into a fully working application in a remarkably short amount of time.

As artificial intelligence becomes increasingly capable of generating written content, the traditional focus on mere word production is no longer sufficient for modern education. Authentic learning must now evolve to prioritize deeper critical thinking and uniquely human skills that technology cannot easily replicate.

This article explores the growing challenge of verifying the quality of AI-generated knowledge work, such as market analyses and professional reports. It questions how professionals can confidently evaluate these outputs without having to manually redo the underlying work to check for accuracy.