Top AI & Tech News Stories — August 10, 2026

Meta Reopens Its Models. Is This a PC Play or a Policy Play?

9 outlets · 12 articles · 15 total since August 10, 2026 · Meta, Muse Glimmer, Mark Zuckerberg, speculative decoding, Hugging Face

The post Meta Reopens Its Models. Is This a PC Play or a Policy Play? appeared first on Futurum . Analyst(s): Nick Patience Publication Date: August 10, 2026 Meta has released Muse Glimmer, a 30-billion-parameter agentic model, under an Apache 2.0 license, and committed to…

More coverage: siliconangle.com · trendingtopics.eu · arstechnica.com · cnbc.com · mlq.ai

Cybersecurity concerns prompt OpenAI to pause some AI training runs

5 outlets · 6 articles · 18 total since August 10, 2026 · Daybreak, OpenAI, Daybreak Blue, incident response, GPT-5.6-Cyber

Cybersecurity concerns prompt OpenAI to pause some AI training runs OpenAI Group PBC recently paused some of its artificial intelligence training workloads over concerns that they could cause cybersecurity issues. The ChatGPT developer disclosed the move in a blog post published…

More coverage: the-decoder.com · cyberscoop.com · cnbc.com · news.google.com

OpenAI Astra AI Model May Reach Critical Cyber Capability Threshold

3 outlets · 3 articles · 70 total since July 21, 2026 · OpenAI, Anthropic, Hugging Face, United Kingdom, incident response

OpenAI has temporarily slowed frontier AI training after internal testing indicated that its upcoming Astra model could approach a critical cybersecurity capability threshold. The decision highlights growing concern that increasingly capable AI systems may independently…

More coverage: thecyberexpress.com · engadget.com

State chatbot laws could create regulatory patchwork, new report warns

1 outlet · 1 article · age verification, content moderation, transparency, parental controls, risk assessment

State chatbot laws could create regulatory patchwork, new report warns As state lawmakers move to regulate artificial intelligence chatbots aimed at children, a report published Monday warns that the growing regulatory patchwork could be difficult for technology companies to…

‘The Curriculum Has Never Been Our Bottleneck. Delivery Is’: Wits Roundtable Reframes AI in African Classrooms

1 outlet · 2 articles · AI governance, responsible adoption, Lithos, DementAI, Richard Meredith

Artificial intelligence will not level African education on its own. It will magnify whatever system it lands in — and the constraint is not what schools teach but whether they can deliver it at all. That was the argument put by Kume Luvhani, co-founder and executive director of…

Old OCR text cripples language model training, and FineBooks wants to fix that at scale

1 outlet · 1 article · Optical Character Recognition (OCR), character error rate (CER), fine-tuning, data reprocessing, Markdown

Old OCR text cripples language model training, and FineBooks wants to fix that at scale Key Points - The FineBooks project, a collaboration between Hugging Face and EleutherAI, benchmarked 14 open-source OCR models on over 2,000 historical book pages to evaluate how well they…

Nigeria Tops Africa on Responsible AI Governance Index — Which Is Not the Same as Deploying AI

1 outlet · 1 article · Responsible AI governance, AI risk management, AI policy development, Digital Sovereignty, data governance

Nigeria has ranked as Africa’s top performer on the 2026 Global Index on Responsible AI, placing 38th of 135 countries with a score of 45.93 — ahead of 72% of countries assessed. The index also makes clear that having policies in place is not the same as having enforceable…

SAMRC and UCT Launch AI Unit Built Around African Health Data, Targeting Cancer, Diabetes, TB and HIV

1 outlet · 1 article · molecular biology, machine learning, bioinformatics, multimodal data integration, pharmacometrics

The South African Medical Research Council and the University of Cape Town have launched an AI for African Population Health Unit that will build machine learning tools trained specifically on African biomedical and clinical datasets — a direct response to the problem that most…

NVIDIA’s Credit Support Buys Exclusivity at OpenAI’s Ohio Data Center

3 outlets · 3 articles · 41 total since December 17, 2025 · Amazon, Anthropic, OpenAI, microsoft, NVIDIA

The post NVIDIA’s Credit Support Buys Exclusivity at OpenAI’s Ohio Data Center appeared first on Futurum . Analyst(s): Brendan Burke Publication Date: August 20, 2026 NVIDIA will provide credit support for the land, power, and shell buildout at SB Energy’s PORTS-Pike Technology…

More coverage: awesomeagents.ai · nytimes.com

Self-flying air taxis and drones: Boeing sells key aviation technology businesses to Archer

1 outlet · 1 article · autonomous flight segments, eVTOL, unmanned aircraft, airspace management, near Vertical Take-Off and Landing (VTOL)

Wisk completed the first flight of its Generation 6 autonomous, all-electric eVTOL last year. Photo via Wisk Archer Aviation is acquiring three Boeing subsidiaries focused on autonomous aircraft, airspace management and unmanned systems, a deal that will significantly expand the…

Transformer Lab wants to automate research with new AI tool Primus

1 outlet · 1 article · Machine Learning (ML), AI agents, large language models, prompt search, model training

Kitchener-Waterloo-based AI research startup Transformer Lab has launched what co-founder and CEO Ali Asaria said is “the most ambitious thing I’ve ever been a part of.” For the past two years, Transformer Lab’s four-person team has been quietly working to fully automate the…

America loves Canadian chips

1 outlet · 1 article · semiconductor startup funding, semiconductor industry, AI strategy, finance automation, Corporate Cards

Exactly 20 years ago, Canadian chipmaker ATI was snapped up by American giant AMD. ATI was, at the time, a dominant player alongside Nvidia—the kind of company that could have anchored a Canadian AI hardware industry. Now, AMD has acquired Toronto-based Taalas—yet another hot…

CISA Warns of Progress LoadMaster Command Injection Vulnerability Exploited in Attacks

1 outlet · 1 article · 2 total since August 8, 2026 · Command injection, Progress, CISA, eSentire, Progress Kemp LoadMaster

CISA has added a critical Progress LoadMaster vulnerability to its Known Exploited Vulnerabilities catalog after attackers were observed targeting exposed devices. Tracked as CVE-2026-8037 , the flaw affects Progress LoadMaster and Progress ADC products. It is a command…

Windows WalletService Vulnerability Allows Attackers to Escalate Privileges – PoC Released

1 outlet · 1 article · privilege escalation, proof-of-concept (PoC) development, DLL loading, Endpoint Detection and Response (EDR), Endpoint telemetry

Microsoft has patched a Windows WalletService vulnerability that could let local attackers gain SYSTEM privileges, with a public proof of concept urging organizations to deploy the July 2026 security updates. Tracked as CVE-2026-49176 , the issue is an elevation-of-privilege…

Apple Private Cloud Compute Flaw Enables Root File Writes and AI Inference Telemetry Leakage

1 outlet · 1 article · path traversal, FAT file system security, cryptographic attestation, secure login, TAR archive extraction

CVE-2026-20685 is a path traversal vulnerability affecting Apple’s Private Cloud Compute (PCC), potentially allowing attackers to write files as root during node boot and redirect sensitive AI inference telemetry to an external server. Sentry Security researcher Drinor received…

ChainDrop worm crawls into npm supply chain, evades standard defenses

1 outlet · 1 article · 2 total since August 10, 2026 · GitHub, npm, malware detection, dependency confusion, typosquatting

A new variant of the Shai-Hulud npm worm has poisoned hundreds of packages while adding propagation techniques that can leave little trace in the corresponding source repositories. In Frank Herbert’s Dune, Shai-Hulud was the name of the giant self-sustaining desert sandworms…

TRAILS Awards More Than $515K for Trustworthy AI Research

1 outlet · 1 article · computer science, Engineering, law, public policy, Social sciences

TRAILS Awards More Than $515K for Trustworthy AI Research Helping students learn to evaluate AI. Making autonomous drones safer. Measuring how people express trust in AI. Helping local governments navigate AI policy. Understanding how to reestablish trust after AI failures…

Red Hat ACM Privilege Escalation Vulnerability Lets Attackers Gain Full Cluster-Admin Access

1 outlet · 1 article · Red Hat Advanced Cluster Management for Kubernetes, privilege escalation, authorization integrity checks, CWE-269, CVSS scoring

Red Hat has disclosed a critical privilege escalation flaw, tracked as CVE-2026-10090, affecting the Application Subscription controller in Red Hat Advanced Cluster Management for Kubernetes (ACM). Rated Important with a CVSS score of 9.9, the vulnerability allows a user with…

Top 10 Malware Threats of the Week – AsyncRAT, Remcos, and Xworm Lead the Surge

1 outlet · 1 article · Remote Access Trojan (RAT), Keylogging, screen capture, data exfiltration, Information-stealing

Global malware activity climbed sharply over the past week, with remote access trojans (RATs), information stealers, and loaders all posting significant week-over-week gains, according to threat sample uploads tracked by ANY.RUN . AsyncRAT topped the chart with 211 uploads…

Valve Steam Hardware Buyers Hit by CEVA Logistics Data Breach

1 outlet · 1 article · cyberattack response, threat detection, Incident investigation, SOC (security operations center), phishing

Valve has confirmed that a cyberattack on CEVA Logistics, its European shipping partner for Steam hardware such as the Steam Deck, Steam Machine, and Steam Controller, exposed customer data belonging to buyers across Europe. The breach occurred between July 29 and August 1…