AI Daily Digest - 2025-10-17

Table of contents

Google DeepMind’s AI Discovers New Cancer Therapy Pathway: A Potential Revolution in Oncology

Category: Healthcare and Life Sciences
Source: Interesting Engineering

In collaboration with Yale University, Google DeepMind has unveiled its groundbreaking 27-billion-parameter AI model “C2S-Scale 27B,” which has uncovered a novel approach to treating cold tumors that evade immune responses. The AI platform identified drugs capable of amplifying immune signaling, with experimental validation proving increased antigen presentation by 50%. This discovery highlights the transformative role AI plays in advancing medical research, offering new hope for patients with difficult-to-treat cancers.

OpenAI Takes Action to Prevent Deepfake Misuse of Historical Figures Like Martin Luther King Jr.

Category: Ethics and Governance
Source: CNBC

Following substantial backlash from Martin Luther King Jr.’s estate, OpenAI has banned its Sora app from generating deepfake videos that depict the civil rights leader. The estate voiced concerns of disrespectful AI representations, prompting OpenAI to tighten its content moderation “guardrails” and give public figures more control over their likenesses in AI-generated media. The controversy shines a spotlight on the ethical dilemmas posed by generative AI’s capabilities.

Major General William ‘Hank’ Taylor Admits Using AI in Military Decisions

Category: Security and Privacy
Source: Dexerto

A top officer in the U.S. Army has shared how he uses ChatGPT to assist in real-time strategic military decision-making. General Taylor noted that AI is increasingly shaping the dynamics of battlefield decisions, which will soon occur “at machine speed” rather than human speed. While praised for its potential, the move has sparked concern over ethical oversight and the risk of sensitive data exposure, highlighting the growing entanglement between AI and national security.

TSMC Becomes the Epicenter of AI Chip Manufacturing Growth

Category: Infrastructure and Hardware
Source: The Next Platform

TSMC reports surging revenue fueled by demand for AI-related chips, with an expected annual revenue surge to over $36 billion in 2025. With recent inroads in 3nm and 4nm processes, the chipmaker is positioning itself as the backbone of AI innovation. NVIDIA and other partners rely heavily on TSMC’s advanced fabrication capabilities for GPUs critical to AI model training and inference.

Rapid Evolution of AI Tools with Claude’s ‘Skills’ and Microsoft’s Windows 11 Copilot Features

Category: Tools and Platforms
Source: Anthropic and Windows Blog

Anthropic introduces “Claude Skills” – specialized task customizations that allow developers to create instructions for dynamic AI responses. Also, Windows 11 rolls out its new Copilot functionality, enabling users to interact with their PC through voice commands while using AI to automate tasks and enhance productivity. These advancements underscore the fast-paced evolution of enterprise AI capabilities in daily applications.

Short News Summary

• Teen Files Lawsuit Against AI Tool Developers: A New Jersey teen sues an AI tool and Telegram for creating fake nude images of her, sparking legal and ethical debates. Read more.
• Rahul Patil Joins Anthropic as CTO: Strengthening Anthropic’s focus on enterprise AI innovation and productivity-forward tools. Read more.
• NVIDIA’s Blackwell Chips Manufactured Domestically: First AI-chip wafers produced in the USA as NVIDIA collaborates with TSMC. Read more.
• California Votes for AI Transparency Law: New AI bills mandate disclosure requirements, but concerns remain over enforcement. Read more.
• ByteDance’s Cici AI Gains International Popularity: Doubao’s sister chatbot gains traction through innovative UI and marketing in global markets. Read more.
• DeepMind Partners with Commonwealth Fusion Systems: AI helps advance plasma control for commercial fusion energy reactors. Read more.