Top AI & Tech News Stories — July 22, 2026
The top AI and tech news story of July 22, 2026 was “OpenAI says its AI agent broke out of testing sandbox to hack Hugging Face”, covered by 34 outlets (48 articles), according to Skillenai's news index. 20 stories ranked across 116 articles.
Last updated · 20 stories · 116 articles
Get the top AI and tech news in your inbox each evening. No account needed, unsubscribe anytime.
OpenAI says its AI agent broke out of testing sandbox to hack Hugging Face
34 outlets · 48 articles · 102 total since July 18, 2026 · Hugging Face, OpenAI, GPT-5.6 Sol, Privilege escalation, cybersecurity
OpenAI says an agent powered by its LLM models escaped its sandboxed testing environment to infiltrate Hugging Face's servers as part of an overzealous attempt to obtain solutions to a benchmark test. The company says it considers the unintended infiltration an "an unprecedented…
Read at news.google.com →First reported by securityonline.infoAMD commits up to $5 billion to Anthropic
9 outlets · 10 articles · AMD, Anthropic, Nvidia, Google, Claude
AMD says it's going to invest up to $5 billion in Anthropic, while helping to expand the AI company's computing power, according to an announcement on Wednesday . As part of the new partnership, Anthropic will deploy up to 2 gigawatts of AMD's Instinct MI450 AI GPUs using the…
Read at platform.theverge.com →First reported by newsroom.amd.comMore coverage: tomshardware.com · cryptobriefing.com · trendingtopics.eu · cnbc.com · edge9.hwupgrade.it
Hugging Face sceglie l’open source cinese Zai_org per difendersi dall’ia che attacca l’ia
18 outlets · 18 articles · 22 total since July 21, 2026 · Hugging Face, OpenAI, GPT-5.6 Sol, ExploitGym, cybersecurity
La vicenda che coinvolge Hugging Face, OpenAI e il modello cinese open weights GLM 5.2 di z.AI apre una delle discussioni più rilevanti del momento sull’evoluzione strategica dell’intelligenza artificiale: la sicurezza dell’AI potrebbe non essere garantita soltanto dai modelli…
Read at tomshardware.com →First reported by xloggs.comMore coverage: transformernews.ai · digitaltrends.com · cnbc.com · news.google.com · bbc.com
Trump tech official accuses China’s Moonshot AI of stealing from Anthropic
5 outlets · 5 articles · 40 total since July 10, 2026 · OpenAI, Anthropic, China, Kimi K3, United States
The Trump administration has accused Chinese start-up Moonshot AI of covertly extracting capabilities from leading US artificial intelligence models and obtaining restricted Nvidia chips abroad, escalating Washington’s scrutiny of the company following the release of its…
Read at scmp.com →First reported by crypto.newsMore coverage: eshumarneedi.com · tcrn.ch · crypto.news · sunmedia.tw
The Army Is Burning Through Its AI Tokens
1 outlet · 1 article · Large language models (LLMs), multimodal generative AI, token-based usage, enterprise subscription E7, Wired
An anonymous reader quotes a report from Wired: A little over a month after the Department of Defense (DOD) bragged that nearly half of its 3.5 million employees were using AI at work, members of the Army's Combat Capabilities Development Command (DEVCOM) received an email…
Jack Dorsey Takes On Slack and GitHub With New AI Workplace Platform 'Buzz'
1 outlet · 1 article · 2 total since July 21, 2026 · Block, buzz, Slack, GitHub, Jack Dorsey
Jack Dorsey's Block has launched Buzz, an open-source workplace collaboration platform that combines messaging, project management, and software development workflows for teams of both humans and AI agents. Dorsey described Buzz as "a new groupchat platform for teams of people…
Read at entertainment.slashdot.org →First reported by thenewstack.ioOpenAI and Anthropic now out-lobby Nvidia in Washington
1 outlet · 1 article · 98 total since March 5, 2026 · OpenAI, United States, Sam Altman, Anthropic, Financial Times
The AI industry is buying a bigger seat at the table. OpenAI and Anthropic set records for federal lobbying in the second quarter of 2026, CNBC reported. Together they spent $3.17 million, up 23% from the first quarter. Startups outspending giants The individual numbers are the…
Read at thenextweb.com →First reported by aidirtylist.infoOracle’s $100 Million Headache: Wisconsin Regulators Stand Firm on Data Center Power Guarantees
1 outlet · 1 article · 5 total since July 21, 2026 · Oracle, OpenAI, Wisconsin, credit rating analysis, S&P
Oracle finds itself staring down more than $100 million in yearly financing expenses. The reason? A nearly one-gigawatt data center campus rising in Port Washington, Wisconsin. Local utility regulators won’t bend on rules meant to shield everyday ratepayers from potential cost…
Read at thenextweb.com →First reported by radiofreemobile.comReddit stock sinks on report it may not renew Google AI content deal
2 outlets · 2 articles · 4 total since July 15, 2026 · Reddit, Google, data licensing, OpenAI, ChatGPT
Shares of Reddit slid 9% on Wednesday after the Wall Street Journal reported the company has discussed shutting off Google's access to its content for artificial intelligence use. Reddit stock is down roughly 27% year-to-date. The two companies struck a deal in 2024 to allow the…
Read at cryptobriefing.com →First reported by digiday.comMore coverage: cnbc.com
Claude launches skill-teaching feature for pro, max, and team plans
1 outlet · 1 article · Claude, Crypto Briefing, Claude, /teach skill
Discover Claude's new 'Teach a Skill' feature for Pro, Max, and Team plans, redefining AI interactions and automating workflows. The post Claude launches skill-teaching feature for pro, max, and team plans appeared first on Crypto Briefing .
IBM and Biji-biji target 41,000 learners with free AI training in Malaysia
1 outlet · 1 article · Artificial Intelligence, agentic AI, software development, web development, AI fundamentals
The program will combine school-level AI education with Agentic AI, software and web development training for TVET learners and adults by the end of 2026. IBM Malaysia and Biji-biji Initiative plan to provide free AI and digital skills training to 41,000 learners by the end of…
After Astronomical Gains, Semiconductor ETFs Fall Back to Earth
2 outlets · 2 articles · 3 total since July 15, 2026 · leveraged exchange-traded funds (ETFs), learning research, equity research, market volatility, options trading
If Sir Isaac Newton had been an ETF investor, his law of gravity would have asserted: What goes up must come down, but will hopefully go right back up again. After an astronomical run for semiconductors in which the memory sector fund DRAM rose almost 200% after launch, the…
Read at insideretail.asia →First reported by congress.netMore coverage: thedailyupside.com
Gli agenti di OpenAI a 10 milioni di utenti settimanali in meno di due settimane
3 outlets · 3 articles · 120 total since July 9, 2026 · OpenAI, ChatGPT, GPT-5.6, ChatGPT Work, codex
Gli agenti di OpenAI a 10 milioni di utenti settimanali in meno di due settimane Il traguardo arriva a meno di due settimane dal lancio di ChatGPT Work, ma è una stima dell'azienda: somma gli utenti attivi settimanali di Codex e ChatGPT Work, non le licenze a pagamento né gli…
Read at apcmag.com →First reported by creativeainews.comMore coverage: edge9.hwupgrade.it · sunmedia.tw
Microsoft And Mistral AI Sign Multi-Billion Deal To Scale Azure Sovereign AI In Europe
8 outlets · 9 articles · 14 total since July 21, 2026 · microsoft, Mistral, Microsoft Foundry, Azure, Nvidia
Microsoft and Mistral AI announced an expanded partnership on July 21, signing a multi-billion dollar deal to scale sovereign Azure capacity across Europe, embed Mistral's models more deeply into Microsoft's product surface and put fresh NVIDIA Vera Rubin GPUs behind the joint…
Read at digitimes.com →First reported by trendingtopics.euMore coverage: punto-informatico.it · theroboticsmedia.com · itmagazine.ch · edge9.hwupgrade.it · corrierecomunicazioni.it
Gafgyt Botnet Hijacks Langflow AI Servers for DDoS Attacks
2 outlets · 2 articles · 63 total since June 30, 2026 · Langflow, Sysdig, JADEPUFFER, MySQL, Nacos
At a Glance Malware family Gafgyt / BASHLITE variant (DDoS bot) Threat actor Unattributed; banner references handle “rimasuta” The post Gafgyt Botnet Hijacks Langflow AI Servers for DDoS Attacks appeared first on Daily CyberSecurity .
Read at thecisobrief.com →First reported by cybersecflux.comMore coverage: securityonline.info
TikTok says AI-generated spam ‘crowds out original creators’ — so it’s teaching users how to spot AI
2 outlets · 2 articles · 40 total since February 1, 2026 · Content moderation, Generative AI, Substack, YouTube, LinkedIn
Graham is the Senior Editor for AI at TechRadar. With over 25 years of experience in both online and print journalism, Graham has worked for various market-leading tech brands including Computeractive, PC Pro, iMore, MacFormat, Mac|Life, Maximum PC, and more. He specializes in…
Read at apcmag.com →First reported by aidirtylist.infoMore coverage: mashable.com
Anthropic wins approval for US$1.5 billion copyright settlement with authors
4 outlets · 4 articles · 115 total since July 5, 2021 · OpenAI, microsoft, ChatGPT, Anthropic, fair use
Anthropic has secured court approval for a US$1.5 billion settlement to resolve a major copyright lawsuit brought by authors over training its Claude chatbot on unauthorized works. The agreement, approved by a US District Court judge, was described as the... The article requires…
Read at digitimes.com →First reported by apfelkraut.orgMore coverage: securityonline.info · news.google.com · edge9.hwupgrade.it
China’s Open AI Models Are Challenging Silicon Valley’s Playbook
2 outlets · 2 articles · 18 total since July 13, 2026 · Anthropic, DeepSeek, China, Google, Z.ai
The AI industry is not quite experiencing a DeepSeek 2.0 moment, but it feels very close. The leading Chinese AI labs have been on a roll lately, releasing a series of almost cutting-edge open-source models. Z.ai released GLM 5.2 in June, Moonshot AI released Kimi K3 last week…
Read at tcrn.ch →First reported by news.800.worksMore coverage: wired.com
Frontier AI accelerates cyber attack development and deployment
2 outlets · 2 articles · 83 total since May 25, 2026 · Anthropic, cybersecurity, Australia, Canada, New Zealand
Essential Insights AI accelerates cyberattacks by enabling less skilled threat actors to craft more convincing phishing, malware, and social engineering campaigns. Attackers exploit AI to automate reconnaissance and bypass traditional security controls, lowering the barrier to…
Read at thecisobrief.com →First reported by laptophub.netMore coverage: securitybrief.ie
Salesforce Announces General Availability of Hyperforce in South Africa
1 outlet · 1 article · cloud platform, data residency, Security compliance, least-privileged access control, Zero Trust principles
Salesforce, the #1 Agentic CRM, today announced that Hyperforce, the company’s trusted cloud platform, is now generally available in South Africa. Delivered through a strategic partnership with Amazon Web Services (AWS), Hyperforce is Salesforce’s premiere infrastructure that…
Get the top AI and tech news in your inbox
One email each evening with the day's biggest stories. No account needed, unsubscribe anytime.
How this was computed
Articles in the Skillenai news index are grouped into event clusters at enrichment time (hybrid embedding/BM25/entity candidate search plus an LLM join-vs-new judge). Stories are ranked by the authority of the articles published on 2026-07-22 (UTC) — the sum of their outlet and author authority — so a multi-day story only ranks on the strength of that day's coverage. Article and outlet counts are likewise scoped to the day. The highlighted link is the day's highest-authority article in each cluster.