Welcome to today’s curated collection of interesting links and insights for 2025/08/29. Our Hand-picked, AI optomized system has processed and summarized 27 articles from all over the internet to bring you the key the latest technology news.
As previously aired🔴LIVE on Clubhouse, Chatter Social, and TikTok.
Also available as a #Podcast on Apple 📻, Spotify🛜, Anghami and Amazon🎧 or anywhere else you listen to podcasts.
1. PassportScan Allegedly Breached -178,000 Scanned IDs for Sale on Dark Web – Daily Dark Web
PassportScan Allegedly Breached -178,000 Scanned IDs for Sale on Dark Web highlights a breach in which 178,000 scanned IDs are allegedly up for sale on the dark web. The description frames this within a broader wave of security threats and database leaks in the cyber underground, noting examples such as unauthorized VPN access and email breaches. It also catalogs multiple related breaches, including Albazaar Shop and the Philippine Civil Service Commission, suggesting a pattern of large datasets being exposed in 2025 and offered for sale. This underscores the ongoing risk to personal data and the need for stronger data protection, proactive monitoring of the dark web, and comprehensive #privacy and #cybersecurity measures in organizations #darkweb.
2. Nvidia’s top two mystery customers made up 39% of the chipmaker’s Q2 revenue
Nvidia’s Q2 results underscore revenue concentration as two customers accounted for 39% of revenue, with Customer A at 23% and Customer B at 16%, up from 14% and 11% a year earlier. The data-center business made up about 88% of total revenue, and large cloud service providers accounted for about 50% of data-center revenue, according to Nvidia’s filing. The two top customers are listed as direct customers, and identities of A and B remain undisclosed, with Nvidia noting it can only estimate revenue to indirect customers, complicating attribution between CSPs, enterprises, and governments. Analysts caution that near-term upside depends on cloud capex clarity, even as demand for Nvidia’s AI systems remains strong across cloud providers, neoclouds, and foreign governments, said @JensenHuang.
3. FBI says China’s Salt Typhoon hacked at least 200 US companies | TechCrunch
Salt Typhoon, a China-backed hacking campaign, has hacked at least 200 US companies and affected organizations in 80 countries, underscoring its global reach. The FBI’s top cyber chief said nine U.S. telecom and internet providers were breached, including AT&T, Verizon, and Lumen, with Charter Communications and Windstream later named as victims. The attackers targeted call records to map who was calling whom, enabling U.S. surveillance under legal orders, and the FBI warned the threat remains ongoing. An international advisory from the FBI offers technical guidance to identify intrusions and mitigate risk, underscoring the ongoing danger to networks, as reported by @ZackWhittaker and highlighting #SaltTyphoon #routers #encryptedMessaging.
@Google Threat Intelligence Group is planning a disruption unit to identify opportunities to take down campaigns, signaling a shift from a purely reactive defense to proactive countermeasures. @SandraJoyce described the effort as intelligence-led proactive identification of opportunities to take down campaigns or operations, emphasizing the need to move from reactive to proactive actions. The article notes ongoing policy debates, including a Center for Cybersecurity Policy and Law report that questions whether the private sector should engage in offensive cyber operations and how deterrence might work. Legislative moves such as the #OneBigBeautifulBillAct and the proposed #ScamFarmsMarqueAndReprisalAuthorizationActof2025 would fund or authorize offensive cyber actions and letters of marque against cybercriminals. Overall, the piece frames a major policy shift under consideration by the US government, with Google signaling readiness to support a more aggressive stance while acknowledging uncertainties about effectiveness and risk.
5. Microsoft asks British Xbox fans to prove their age
Microsoft is introducing age verification for Microsoft accounts in the UK as part of compliance with the UK Online Safety Act. The company is emailing UK Xbox users that verification will be required for access to Xbox experiences, with options including #age-verification via age estimation, ID checks, mobile contract verification, and credit card verification, delivered through @Yoti. Microsoft says the goal is to provide age-appropriate experiences and keep the Xbox community safe, even as the act’s enforcement has prompted ways to bypass checks such as VPNs and even AI-generated documents. @Dame_Rachel_de_Souza has suggested adding VPN providers to the list of age-gated services, highlighting regulatory pressure and enforcement tensions. The move shows how platforms are applying #UKOnlineSafetyAct requirements with multi-layer verification, reflecting the tension between safety objectives and practical evasion.
7. Microsoft AI launches its first in-house models
Microsoft unveils its first homegrown #AI models, MAI-Voice-1 and MAI-1-preview, signaling a shift toward consumer-oriented capabilities within its #Copilot ecosystem. MAI-Voice-1 can generate a minute of audio in under one second on a single GPU, while MAI-1-preview offers a glimpse of future Copilot offerings and is already used to power features like Copilot Daily and podcast-style discussions. Microsoft trained MAI-1-preview on about 15,000 #NvidiaH100 GPUs and plans to roll it out for certain text use cases in Copilot, with public testing on the AI benchmarking platform #LMArena. As @MustafaSuleyman notes, the focus is on consumer utility rather than enterprise use cases, reflecting a strategy to optimize for consumer data and experiences and to orchestrate a range of specialized models serving different intents to unlock value. This move complements Microsoft’s broader AI ambitions and its evolving partnership with @OpenAI as it pursues a multi-model approach beyond relying solely on external LLMs.
8. Trump Nixes Patent Office, Weather Service, NASA Unions (1)
President @DonaldTrump expands a March executive order to end collective bargaining at multiple federal agencies, labeling them as #NationalSecurity interests and exempting them from #FederalUnionLaws. The directive terminates collective bargaining agreements at #NASA, the #InternationalTradeAdministration, the #PatentOffice, the #NationalWeatherService, the #USAgencyforGlobalMedia, hydropower facilities under the #BureauofReclamation, and the #NESDIS program, arguing that labor-management delays can hinder swift policy implementation for critical missions. This move follows a Supreme Court victory that allowed this approach in some agencies and continues Trump’s broader effort to weaken the career civil service and blur the lines between politics and day-to-day governance, a shift critics say raises questions about presidential power. Already-cancelled contracts at Health and Human Services, Veterans Affairs, Homeland Security, and the EPA illustrate the expansion of this policy and the White House’s claim that faster decision-making is needed to protect national security.
9. Twitch denies viewership “free fall” and rumors of viewbot update rollback – Dexerto
While rumors circulated about a viewership free fall on Twitch, the company says there has been no such decline and that it has not rolled back its viewbot updates. In July 2025, Twitch announced changes to better identify viewbots and inauthentic engagement, with CEO @DanClancy noting the updates are being implemented carefully to avoid filtering real users. Data from third-party trackers such as @TwitchTracker posted a downtrend since Aug 21, but Twitch maintains those numbers are not Twitch data and are inaccurate. Twitch emphasizes that viewership metrics on its platform should reflect real communities and that combatting viewbots remains a long-term priority, not a short-term fix. The article notes that Kick has faced similar issues, with prominent streamers like @xQc noting that bot-detection has advanced, underscoring the wider botting challenge.
10. Trump picks Jim O’Neill to head CDC foundation
Former President @Donald Trump selected Jim O’Neill to lead the CDC Foundation, reflecting a shift in the agency’s external partnerships amid ongoing public health challenges. O’Neill, known for his experience in public-private collaborations, is expected to focus on enhancing innovation and resource mobilization for the CDC. His appointment aims to strengthen the Foundation’s role in supporting the CDC’s mission in disease prevention and emergency response. This move highlights the ongoing efforts to adapt the CDC’s operational strategies to better address contemporary health threats. The selection underscores the importance of integrating diverse expertise to bolster national health security.
11. Tesla FSD turns off more U.S. consumers than its attracts, survey finds
A new poll finds U.S. consumers are more likely to be turned away from @Tesla by its #FSD (Self-Driving) than drawn toward it. An August EV Intelligence Report from @SlingshotStrategies, which surveyed 8,000 Americans, shows only 14% would be more likely to buy a Tesla, 35% would be less likely, and 51% said the feature would make no difference. Researchers note reputational damage tied to @ElonMusk’s rhetoric and political associations, including work with the Trump administration and support of #AfD, as a potential drag on demand. The findings come as Tesla faces a sales slump, a lagging #robotaxi effort behind #Waymo and #Baidu, and a still-evolving autonomous strategy as it tests rides in Austin and the San Francisco Bay Area. Analysts say broader education and demonstrated, reliable autonomous performance are needed for FSD to help Tesla regain momentum, especially with a new Model Y variant delaying production in the U.S. until late next year.
12. Intel gets $5.7 billion from Trump deal as White House says details are ‘being ironed out’
Intel received $5.7 billion in cash from the U.S. government as part of the White House’s decision to take a 10% stake in the company, with details still being ironed out by the Department of Commerce. CFO @David Zinser disclosed the cash infusion during an investor conference, and Intel signaled a potential opportunity for outside investors in its #foundry business. The company had better-than-expected second-quarter results, but shares fell about 8% on concerns about the foundry unit’s business prospects. The corporate filing warned of possible adverse reactions from investors, employees, customers, suppliers, foreign governments or competitors and potential litigation, highlighting ongoing public and political scrutiny of the transaction.
13. Vivaldi takes a stand: keep browsing human | Vivaldi Browser
Vivaldi declares browsing should be human-centered and active, not a gateway to passive consumption. Across the industry, AI-assisted browsing is rising as Google adds Gemini to Chrome to summarize pages and potentially navigate across tabs, while Microsoft markets Edge as an AI browser with modes that scan on screen and anticipate actions. Independent research shows AI summaries can reduce engagement with original sources, with Pew Research finding that traditional results are clicked roughly half as often when AI summaries appear. The next phase of the browser wars centers on who intermediates knowledge, who benefits from attention, who controls the information pathway, and who monetizes users. Vivaldi commits not to embed an LLM for chat, summarization, or auto-fill until more rigorous privacy-preserving methods exist, and will use AI only if it advances exploration without stealing IP or compromising privacy, as @Jon von Tetzchner, CEO, states that they will stay true to a powerful personal browser for curious minds and autonomy. #Vivaldi #openweb #AI #privacy
14. Trump Tariffs Cause Chaos on eBay, Having a Hobby Is Now a Logistical Nightmare
The imposition of tariffs under @realDonaldTrump’s administration has created significant disruptions for eBay sellers and hobbyists by increasing costs and complicating shipping logistics. Many users report unexpectedly high fees and delays in receiving supplies or sending products internationally, making formerly simple transactions cumbersome. This new environment forces individuals to reassess their buying and selling habits, often discouraging participation in hobby-related markets. The changes illustrate how #trade policies can ripple through everyday activities, affecting not just businesses but also casual consumers engaged in hobbies. Consequently, the tariffs have transformed the landscape of online markets, turning what used to be straightforward into a logistical challenge.
15. Blogging service TypePad is shutting down and taking all blog content with it
TypePad, once a user-friendly blogging platform built on @MovableType, will shut down on September 30, taking all posts and hosted content with it, leaving current users a little over a month to export their data. The service stopped accepting new accounts in 2020, and executives gave no specific rationale beyond a difficult decision, despite earlier assurances that there were no plans to shut down as recently as March. TypePad’s shutdown continues a trend of aging internet services shrinking or disappearing, with content likely limited to the Internet Archive’s #WaybackMachine. The history involves @SixApart, @SayMedia, @InfoCom, and @EnduranceInternationalGroup, illustrating how consolidation and corporate ownership have shaped legacy #CMS and hosting brands, while @MovableType remains active as a separate product. The move underscores how keeping old content accessible depends on external archives and the fragility of once-dominant platforms in the evolving web landscape #Web2.0.
16. AI Doesn’t Spout MAGA Propaganda. No Wonder Donald Trump Hates It.
The article argues that @Trump’s July executive order to prevent #WokeAI in the federal government reveals his movement’s claim that reality has a liberal bias that must be policed by policy. It notes that models like Claude provide nuanced, apolitical discussions of gender and diversity, while the order insists such topics threaten ‘truth’ and reframes diversity guidelines as a danger to factuality. The piece explains that #LLMs learn from massive datasets and cannot be constrained to Trump’s edicts without sacrificing utility, illustrating the gulf between his demands and how #LLMs actually function. It argues that Trump misreads social media as the internet’s killer app and ignores AI’s practical uses, since most users turn to AI for coding, navigation, weather, and other everyday tasks rather than fear-driven content. By portraying the policy as a DARVO-like manipulation that weaponizes #diversity and #fairness claims as threats to truth, the piece claims reality resists political control of #AI and that Trump’s approach would distort how AI serves the public.
17. Taiwan prosecutors indict three people for stealing TSMC trade secrets to help China’s SMIC
Taiwanese prosecutors have indicted three individuals for allegedly stealing trade secrets from @TSMC, the world’s largest contract chipmaker, to assist China’s #SMIC in developing 7-nanometre chip technology. The suspects reportedly worked for a Taiwanese supplier and transferred confidential information to #SMIC, a move that breaches national security and harms Taiwan’s semiconductor industry. The indictment highlights ongoing concerns about industrial espionage amid US-China tech rivalry, as Taiwan seeks to protect its technological advances. This case underscores the challenges Taiwan faces in safeguarding its leading position in semiconductor manufacturing against espionage threats.
18. Some teachers are using AI to grade their students, Anthropic finds – why that matters
Some educators are using @Claude to help grade students, a controversial use that reveals educators’ adoption of #AI in the classroom extends into assessment. Anthropic’s Education Report analyzed 74,000 anonymized Claude conversations from May and June 2025 among higher-ed accounts and mapped each to #O*NET education tasks. Findings show #curriculumDevelopment (57%) and #academicResearch (13%) as the most common AI uses, with 7% of educators using Claude to assess student performance, including feedback, grading against rubrics, and summarizing evaluations. When used to grade, AI automation occurred nearly half the time (48.9%), despite concerns about automating assessment and AI’s perceived limited effectiveness in this area. The pattern suggests educators favor AI as a supporting copilot for instruction, grant proposals, advising, and administrative chores, while treating grading as a more debated application.
In a blind comparison of eight very short fantasy stories, four AI-written and four by renowned authors including @MarkLawrence, @JannyWurts, @ChristianCameron, and @RobinHobb, readers could not reliably tell which pieces were AI and which were human, and AI-written stories were rated higher on average. The test on Lawrence’s blog drew 964 voters who correctly identified three stories, misidentified three, and could not decide on two, with the AI-written piece finishing as the top-rated story. Lawrence cautions that shorter outputs are where AI tends to sound most human and that the eight stories were produced by novelists who typically write long-form work, so the result may not reflect flash-fiction proficiency; beyond the quiz, AI-generated entries have flooded sci-fi magazines, prompting one publication to halt submissions, and platforms like Steam and Kindle now rely on self-reported AI elements, which may be unsustainable. The piece argues that trying to police AI art by instinct alone is unreliable and calls for more rigorous verification and clearer standards as #AIwriting continues to spread and challenge traditional authors.
20. World’s most powerful solar telescope sees incredible coronal loops on the sun (image)
New observations from the Daniel K. Inouye Solar Telescope, DKIST, captured hundreds of coronal loop strands averaging 29.95 miles (48.2 km) in width, with some as thin as 13 miles (21 km), imaged in hydrogen-alpha light after an Aug. 8, 2024 X-class flare #coronal_loops #X_class_flare. Its resolution is more than 2.5 times sharper than the next best solar telescope. As @Tamburri notes, this enables researchers to move from seeing a forest of loops to resolving individual loops, which may be fundamental building blocks of the sun’s magnetic architecture that drive flares #magnetic_reconnection. Researchers are still working out how these small loops participate in magnetic reconnection, but the new detail helps integrate them into models of solar activity. Budget proposals for FY 2026 would cut DKIST funding from $30 million to $13 million, risking closure and the loss of data and training opportunities.
21. World’s first method turns plastic into fuel with 95% efficiency
A #one-step method converts mixed plastic waste into petrol at #room_temperature and #ambient_pressure, achieving over 95% efficiency. The outputs include the main components of #gasoline, plus chemical raw materials and hydrochloric acid, enabling feeds into #water_treatment, #metal_processing, #pharmaceuticals, #food_production, and the #petroleum_industry. The authors say the approach requires less energy, less equipment, and fewer steps than conventional plastic-to-fuel pathways, making it scalable for industrial use #less_energy #less_equipment #fewer_steps. The work involves researchers from @PNNL, @Columbia_University, @TechnischeUniversitaet_Muenchen, and @ECNU, and supports a #circular_economy by turning diverse plastic waste into valuable products in a single step.
Russian technology company @Yandex was fined for declining to provide the Federal Security Service (#FSB) with 24/7 access to its smart home system. The fine stems from the company’s refusal to comply with government demands for constant surveillance capabilities. This incident highlights the ongoing tension in Russia between tech corporations and state security agencies over privacy and control of digital technologies. The case exemplifies the challenges companies face in balancing user privacy with governmental pressure in the Russian regulatory environment. It reflects broader debates on state influence over #smarttechnology and digital security in Russia.
25. SpaceX launches record-breaking 30th flight of a Falcon 9 booster – Spaceflight Now
A @SpaceX #Falcon9 booster, tail number B1067, marked its 30th flight as it lifted the Starlink 10-11 mission from Launch Complex 39A at NASA’s Kennedy Space Center. Liftoff from pad 39A occurred at 4:12 a.m. EDT and SpaceX confirmed the deployment of 28 #StarlinkV2Mini satellites into low Earth orbit. Its prior missions include two astronaut flights, two cargo missions to the #ISS and 18 batches of Starlink satellites, underscoring SpaceX’s reuse and launch cadence in 2025 across 74 missions. Nearly 8.5 minutes after liftoff, B1067 was set to land on the drone ship ‘A Shortfall of Gravitas’ in the Atlantic, with the landing expected to be the 122nd on that vessel and the 495th booster landing to date.
NASA has opened a new Mission Evaluation Room at @NASA’s Johnson Space Center to support the Artemis 2 Orion mission, complementing the main WFCR in mission control. The MER features 24 console stations and will be staffed 24/7 for roughly 10 days, with personnel from @LockheedMartin, @ESA, and @Airbus responsible for different parts of the spacecraft. MER will monitor #Orion and its many systems, compare real-time data with expected performance, and troubleshoot issues, serving as Orion’s engineering brain trust and enabling reachback support from an international team. Beyond real-time troubleshooting, MER will collect detailed mission data to inform Artemis 3 planning and future missions, underscoring a leap in technology and collaboration among international and private-sector partners. With MER, NASA’s Mission Control becomes effectively a two-room operation, one to fly Orion and one to monitor its health, setting a standard for Artemis 2 and guiding Artemis 3 and beyond toward a sustained lunar presence.
27. NASA Seeks Volunteers to Track Artemis II Mission – NASA
NASA is seeking volunteers to passively track the Artemis II Orion spacecraft to inform a shift toward a commercial-first approach to space tracking. The crewed mission, targeting no later than April 2026, will launch the SLS rocket with astronauts @ReidWiseman, @VictorGlover, @ChristinaKoch and CSA @JeremyHansen on a roughly 10 day loop around the Moon before returning to Earth, and it will rely on NASA’s #NearSpaceNetwork and #DeepSpaceNetwork for primary communications and tracking. Past experience from Artemis I showed a diverse mix of participants who attempted to receive Orion signals using their own ground antennas, helping NASA measure radio wave changes and inform capabilities outside government. This data will inform strategies to broaden tracking, strengthen infrastructure for Artemis missions #MoonToMars objectives, and support a future commercial-led backbone for space communications and navigation through the SCaN program. By expanding industry participation, NASA aims to validate #commercial-first approaches and ensure robust support for human deep space exploration.
That’s all for today’s digest for 2025/08/29! We picked, and processed 27 Articles. Stay tuned for tomorrow’s collection of insights and discoveries.
Thanks Patricia Zougheib and Dr Badawi for curating the links
See you in the next one! 🚀