Digimasters Shorts - Apple sued for pirated AI training books, OpenAI escapes permanent data logging, Trump deploys troops over AI-faked riots, MIT advances nuclear fusion control, Microsoft integrates AI Copilot with Google apps

Digimasters Shorts is your quick-hit source for the latest developments in the digital world. Hosted by Adam Nagus and Carly Wilson, this podcast delivers concise updates on AI legal battles, breakthroughs in fusion energy, technological innovations, and societal impacts shaping our digital landscape. Whether it's the ongoing legal challenges faced by AI giants like Apple and OpenAI, cutting-edge advancements in nuclear fusion research at MIT, or new features in Windows that blend Google and Microsoft tools, Digimasters Shorts keeps you informed in a fast, engaging format. Stay ahead of the curve with expert insights and breaking stories, all in under a few minutes.

Transcript

Adam N2: 0:04

Welcome to Digimasters Shorts, we are your hosts Adam Naygus

Carly W: 0:08

and Carly Wilson delivering the latest scoop from the digital realm. Apple is facing a new class action lawsuit accusing the company of using pirated books to train its AI models. Neuroscience professors Susana Martinez-Conde and Stephen Macknik from SUNY Downstate Health Sciences University allege Apple used their registered works without authorization. The plaintiffs claim Apple employed"shadow libraries" and web-crawling software to access pirated copyrighted books, including their own. This lawsuit follows a similar case where another pair of authors accused Apple of copyright infringement in AI training. Apple is not alone in these legal challenges; Open A.I has also been sued by The New York Times over alleged unauthorized use of published works. These cases highlight growing concerns about copyright violations in AI training materials. Earlier this year, Anthropic settled a major class action lawsuit by paying$1.5 billion to 500,000 authors. That settlement marks a significant precedent in copyright lawsuits related to AI. The technology’s rapid advancement has brought intellectual property issues to the forefront. The outcomes of these suits will likely shape the future of copyright law in AI development.

Adam N2: 1:20

A federal judge has lifted the court order requiring Open A.I to preserve all Chat G.P.T data indefinitely. The original order, issued in May, was part of a lawsuit by the New York Times claiming Open A.I trained its AI on the paper's content without permission. Several other publishers, including The Intercept and Ziff Davis, have also sued Open A.I and Microsoft for copyright infringement. The May order forced Open A.I to keep logs of Chat G.P.T outputs, which the New York Times used to investigate copyright claims. Open A.I opposed the order, citing concerns over user privacy and data security, but initially lost the fight. Judge Ona T. Wang ruled that Chat G.P.T users were non-parties to the lawsuit, allowing plaintiffs access to preserved logs. The new order with a few exceptions ends Open A.I's obligation to maintain logs past September 26. However, any logs saved during the preservation period remain accessible to the plaintiffs. Open A.I must still retain data linked to accounts flagged by the New York Times. This development marks a partial victory for Open A.I amid ongoing legal battles over AI training data. President Donald Trump has claimed that American cities are overrun with riots and anarchist violence, prompting him to deploy federal troops to Portland, Washington DC, and Memphis. However, investigations reveal a lack of concrete evidence supporting widespread violent protests attributed to left-wing groups. This gap has been filled by AI-generated videos, created using tools like Open A.I's Sora 2, which produce photorealistic clips depicting violent confrontations between protesters dressed in black and militarized officers. These videos have gone viral on social media, often without viewers recognizing their artificial nature. The clips reinforce stereotypes of aggressive protestors based on black bloc tactics, helping to justify calls for a crackdown labeled by Trump as a"war." Despite these portrayals, crime statistics show violent crime in the U.S. is at multi-decade lows, and the supposed anarchist uprisings are largely absent. After troops were deployed to DC, they quickly exhausted their targets and were reassigned to tasks like cleaning streets at a significant cost. This use of AI-generated propaganda raises concerns about manipulating public perception to support costly government interventions. The phenomenon highlights the role emerging technologies play in shaping political narratives and governance. How long this campaign of manufactured crisis will persist remains uncertain.

Carly W: 3:50

Researchers at M.I.T have made progress in managing plasma behavior inside tokamak reactors, a key step toward achieving practical nuclear fusion. Fusion promises a clean, abundant energy source by replicating processes powering stars. Tokamaks use strong magnets to confine plasma at temperatures hotter than the Sun’s core, but safely controlling and shutting down these reactions remains challenging. The M.I.T team combined physics principles with machine learning to predict plasma behavior during reactor rampdowns, reducing risks of damage. Their model was trained using data from Switzerland’s TCV fusion device, enabling safe guidance for plasma de-energization. Applying the algorithm, operators improved shutdown procedures, minimizing wear on reactor components. Lead researcher Allen Wang emphasized that reliable plasma control is essential for fusion’s practical use. This breakthrough addresses a significant barrier in fusion technology, making reactors more manageable and potentially closer to commercial viability. The research marks a promising advancement, although scientists acknowledge fusion still faces a long road ahead. Their work opens new pathways for safer, more efficient nuclear fusion energy. Microsoft has introduced a new Windows update that integrates its Copilot AI with Google apps, allowing users to perform cross-app searches and tasks. This update supports Google Drive, Gmail, Google Calendar, and Google Contacts alongside Microsoft Outlook and OneDrive. Copilot, Microsoft's generative AI rivaling Google Gemini and Chat G.P.T, can now search through users’ files, emails, and contacts with voice or text commands. The integration operates via Microsoft Connectors and requires users to opt in through the Windows settings menu. Users can enable Copilot individually for each supported app, giving them control over which data Copilot can access. Examples include asking for specific email addresses or retrieving notes from connected Google accounts. Version 1.25095.161.0 and higher began rolling out to Windows Insiders on October 9, 2025, with a gradual release to all users. Those not yet seeing the feature after updating are advised to wait for it to appear. This update marks a rare collaboration between Microsoft and Google and aims to enhance productivity for Windows users. Opt-in activation ensures user privacy by preventing unauthorized AI access to data.

Don: 6:16

Thank you for listening to today's AI and Tech News podcast summary... Please do leave us a comment and for additional feedback, please email us at podcast@digimasters.co.uk You can now follow us on Instagram and Threads by searching for@DigimastersShorts or Search for Digimasters on Linkedin. Be sure to tune in tomorrow and don't forget to follow or subscribe!