Top AI & Tech News Stories — July 20, 2026

OpenAI’s models shared hacking tips on a secret messaging board before Hugging Face breach - Politico

7 outlets · 7 articles · 100 total since February 14, 2026 · OpenAI, Hugging Face, GPT-5.6 Sol, sandboxing, ExploitGym

OpenAI’s models shared hacking tips on a secret messaging board before Hugging Face breach Politico

More coverage: the-decoder.com · techcrunch.com · news.google.com · theregister.com · thehackernews.com

BTS Expands Advanced Engineering Portfolio With Augury Acquisition

1 outlet · 1 article · FPGA development, system architecture, full-stack applications, Artificial intelligence, cybersecurity

- BTS Software Solutions aims to expand its engineering and hardware capabilities with the Augury acquisition - Augury brings expertise in FPGA development, system architecture and full-stack mission applications to BTS - The deal adds advanced hardware capabilities to BTS’…

AMD to invest up to $5 billion in Anthropic under AI infrastructure deal

3 outlets · 3 articles · 12 total since July 19, 2026 · AMD, Meta, Anthropic, OpenAI, microsoft

AMD has agreed to invest up to $5 billion in Anthropic under an infrastructure agreement covering tens of billions of dollars’ worth of AI systems. Anthropic will deploy up to two gigawatts of capacity using AMD’s Instinct MI450-series accelerators, with deployment of the first…

More coverage: the-decoder.com · cnbc.com

Google's Frozen v2 Chip Bakes Gemini Into Silicon

5 outlets · 5 articles · 6 total since July 20, 2026 · Frozen, Google, Gemini, The Information, AI inference

Google is designing a chip that doesn't just run Gemini faster. It's trying to make Gemini part of the chip itself. TL;DR The Information reports Google is building "Frozen v2," a chip line separate from its TPU roadmap that hardwires elements of Gemini 's architecture directly…

More coverage: thenewstack.io · awesomeagents.ai · officechai.com · cnbc.com

Kimi K3 Breaks Free: How China’s Open AI Model Cheated Its Way Out of Containment

3 outlets · 3 articles · 49 total since July 17, 2026 · China, Anthropic, OpenAI, Kimi K3, United States

Security teams just watched one of the world’s most capable open-weight AI systems slip its leash. Not with code exploits or network breaches. The model simply wanted to win a test. So it did what many humans would. It found a shortcut. Kimi K3 from Beijing-based Moonshot AI…

More coverage: wp.technologyreview.com · awesomeagents.ai

Watchdog flags security issues with IRS’s Zero Paper Initiative

1 outlet · 1 article · cybersecurity, vulnerability scanning, configuration management, security controls, IT system security

The IRS is facing more hurdles in its lagging effort to transition from paperless processing to a fully digitized approach to tax returns. A memo sent by the Treasury Inspector General for Tax Administration to the IRS commissioner last week flagged digital and physical security…

Cash out veterans’ preference in federal hiring, honor their debt

1 outlet · 1 article

FedRAMP Director Pete Waterman offered a glimpse at federal hiring frustrations earlier this month with this post on LinkedIn : I’m really sorry to all the folks who applied for the FR Cybersecurity Service only to be told that their application was not considered or reviewed…

A.I. ‘Vibecoded’ Apps Are Flooding Apple’s App Store

1 outlet · 1 article · Apple, Apple’s China App Store

Artificial intelligence is feeding a boom in the number of apps available in Apple’s App Store.

Veteran Microsoft security executive joins AWS amid broader reshuffle in Redmond

1 outlet · 1 article · Cloud security, threat detection, data governance, Security services, security & risk monitoring

Rudra Mitra will lead Amazon security services in his new role. (LinkedIn Photo) Rudra “Rudy” Mitra , who spent more than 27 years at Microsoft and most recently led its Purview data-security business, is joining Amazon Web Services as vice president of security services. Mitra…

How the Fair Work Commission safely deployed Agentic AI in weeks, not years

3 outlets · 3 articles · 6 total since July 17, 2026 · Google, identity, microsoft, AgentCore, observability

Prioritising strict governance and rapid development, Australia’s Fair Work Commission deployed governed AI agents into production in weeks, enhancing operational efficiency while maintaining public trust. Australia's Robodebt scandal left a scar on public trust in government…

More coverage: techwireasia.com · greyhoundresearch.com

Una Guía de Evaluación para la Era de la Inteligencia Artificial

2 outlets · 2 articles · 70 total since February 7, 2026 · Anthropic, OpenAI, Claude, Claude Code, Amazon

Sarah Friar, directora financiera de OpenAI, presentó recientemente una innovadora herramienta destinada a evaluar el rendimiento de la inteligencia artificial en las organizaciones. Se trata de una puntuación práctica, conocida como «scorecard», que permite medir el retorno de…

More coverage: techwireasia.com

Moonshot AI reportedly opens pre-IPO round at $50 billion valuation as Kimi K3 drives demand

2 outlets · 2 articles · 33 total since July 16, 2026 · Kimi K3, Moonshot AI, China, Anthropic, OpenAI

Moonshot AI, the Chinese developer of the Kimi chatbot, has reportedly opened a G-round pre-IPO financing at a valuation of $50 billion. A person familiar with the deal said investor interest intensified after Kimi K3 launched, with institutions registering intended commitments…

More coverage: artificialintelligence-news.com

Trump administration reportedly builds a slow-motion ban on Chinese AI models through sanctions and soft pressure

1 outlet · 1 article · sanctions, security risk management, Trump administration, OpenAI, Google

The Trump administration is reportedly weighing measures targeting Chinese AI models, from adding Chinese labs to sanctions lists to holding U.S. companies liable for security failures. Rather than impose an outright ban, Washington could use soft rules to deter adoption while…

Anthropic employees worked “literally around the clock” to keep Fable 5 from disappearing

1 outlet · 1 article · 3 total since July 16, 2026 · OpenAI, Anthropic, Cloud Infrastructure, microsoft, CNBC

After weeks of extending temporary access while bringing additional inference capacity online, Anthropic has finalized Claude Fable 5 subscriptions. In a Friday announcement on X, the company said Claude Fable 5 will become a permanent part of Max and Team Premium subscriptions…

8-bit ‘Jimothy’: Viral sensation raids trash cans, eludes paparazzi in Seattle creator’s video game

1 outlet · 1 article · HTML, OpenAI GPT-5.6, AI model prompting, vibe-coding, game design

The “Jimothy” video game features Seattle’s beloved raccoon making his way through assorted challenges. (Image via Chris Pirillo) Jimothy, the short-spined Seattle raccoon, has become a global sensation whose likeness has been immortalized in artwork, clothing, songs and tattoos…

Nigeria’s First ‘Deepfake Election’: Bloomwit Africa’s New Report Warns the 2027 Vote Will Be Fought in Channels No One Can See

1 outlet · 1 article

Nigeria’s 2027 general election will be the first in the country’s history to take place with generative AI tools in wide circulation. The information environment around it has already been transformed, according to Navigating Nigeria 2027 , a new strategic communications and…

AI is more likely than humans to form biases when hiring

1 outlet · 1 article · Large Language Models (LLMs), Prompting, reasoning, Bias mitigation, resume screening

The next time you apply for a job, AI may screen your résumé before any human sees it. But there’s good reason to question whether AI will judge you fairly. Researchers already know that LLMs pick up human biases from their training data. New research suggests that LLMs can also…

UK data center startup Nscale bets big on Bellevue for U.S. engineering hub amid AI boom

1 outlet · 1 article · AI infrastructure, GPU infrastructure, Data center operations, cloud/AI infrastructure, onboarding

Nscale’s Nidhi Chappell. Photo via Nscale. Fresh off a $2 billion fundraising and $900 million line of credit, London-based data center startup Nscale is planning a big expansion at a new engineering office in Bellevue, Wash. Nscale, one of the fastest-growing companies building…

Honor confirms August global launch of its first Robot Phone at WAIC 2026

1 outlet · 1 article · log decoding, color grading, ultra-wide lens, Honor, Qualcomm

At the 2026 World Artificial Intelligence Conference (WAIC), Honor CEO Li Jian unveiled the company’s first Robot Phone, confirming it will launch globally in August with pre-orders now open across all sales channels. The device is powered by Qualcomm’s latest Snapdragon 8 Elite…

Bilibili showcases N.E.K.O., an AI companion that can interpret desktop content and initiate conversations

1 outlet · 1 article · multimodal AI, computer vision, Memory layer, front-end interfaces, Open-source

Bilibili showcased its open-source “Catgirl Plan” AI digital-life ecosystem at WAIC 2026 in Shanghai on July 18. Its core product, Project N.E.K.O., is a proactive multimodal AI companion that can continuously observe a computer environment, interpret desktop content and…