Wazafak — AI Business

Meta's digital signature has been found to be unstable, with a high probability of matching random images. This raises concerns about the security and reliability of Meta's systems. The company's use of artificial intelligence and immersive technologies has also been called into question.

GoPro's financial struggles may lead to its demise, with founder Nicholas Woodman loaning $20 million to keep the company afloat. The action camera specialist has experienced a 26% year-on-year revenue drop, and its future looks uncertain.

A group of 25 tech companies, including Nvidia, Microsoft, and Meta, have released a letter urging policymakers to avoid 'premature restrictions' on open-weight AI models, citing concerns that such restrictions could stifle competition and drive innovation overseas.

Claude Opus 5 is a new AI model that approaches frontier intelligence at half the price of Fable 5. It excels in software engineering tasks and is the new default model on Claude Max. However, it remains behind Mythos 5 in cybersecurity tasks and biology research.

The Islamic Revolutionary Guard Corps claims to have destroyed Amazon's data center in Bahrain with cruise missiles, in retaliation for the US strike on Iran's Darkhovin nuclear power plant. This marks the first use of cruise missiles against commercial cloud infrastructure in the Gulf conflict.

Harness engineering is not enough to ensure software factories succeed, as it fails to address the fundamental issue of model training and codebase maintainability.

AMD and Cerebras join forces to develop a disaggregated compute platform, combining Instinct GPUs with Cerebras' SRAM-powered AI accelerators to deliver ultra-low-latency inference for AI workloads, challenging Nvidia's Groq LPUs.

Nearly 200 Silicon Valley companies are urging the Trump administration not to cut off access to Chinese open-weight AI models, warning it could cripple the next generation of U.S. startups. The founders argue that American leadership in AI innovation requires access to both domestic and global open models.

Five US tech giants are hiding an estimated $1.65 trillion in debt that doesn't appear on balance sheets, sparking concerns of an AI bubble. The companies, including Alphabet, Microsoft, Amazon, Meta, and Oracle, are using special purpose vehicles to keep the debt off their official financial reports.

Alphabet's first cash burn on record has jolted investors awaiting more Big Tech results next week as soaring AI spending strains one of the world's most profitable companies.

LG Electronics USA announced plans to suspend apps that turn smart TVs into residential proxy nodes, following research that found 42% of games and apps on LG's webOS store allow unknown third-parties to route internet traffic through users' TVs.

A US court has ruled in favor of Apple, dismissing a lawsuit that claimed the company was liable for not scanning its iCloud storage for child sexual abuse material (CSAM). The case highlights the tension between privacy and the need to prevent the spread of CSAM.

The EU Court has ruled that VPNs are lawful technical tools, shielding publishers and VPN providers from liability in copyright disputes, even when users bypass geo-blocking measures.

France's national cybersecurity agency, ANSSI, will stop certifying security products lacking quantum-resistant encryption starting 2027, accelerating the country's transition to post-quantum cryptography. This move will effectively require government agencies and operators of critical infrastructure to move away from traditional cryptographic systems.

Jack Dorsey's new platform, Buzz, combines team chat, AI agents, and Git hosting, aiming to reduce dependence on Slack and GitHub. The open-source workspace uses signed Nostr events and a self-hostable relay architecture.

The Electronic Frontier Foundation (EFF) and other public interest organizations have urged the Federal Trade Commission (FTC) to reject X Corp's petition to modify its consent order, citing concerns over the company's data handling practices and potential risks to consumer privacy.

Madison Square Garden compiled a dossier on activists who opposed its facial recognition program, raising concerns about surveillance and biometric data collection. The document, which included tweets and comments from critics, was accessible to MSG staff and has sparked concerns about the venue's use of surveillance technology.

A hacker has breached Romania's cadastre agency and wiped the country's entire land registry database, bringing the real-estate market to a standstill. The hack has left notaries unable to record new transactions and citizens unable to obtain proof of ownership or detailed land records.

The UK government's Matrix project, a shared services cluster strategy, has been rated red by the National Infrastructure and Service Transformation Authority, indicating major issues with its schedule, budget, quality, or delivery of benefits. The project, which aims to move nine government departments to shared ERP and HR systems, has experienced delays and is now considered unachievable without urgent action.

The current AI mania is distorting corporate strategy, with executives making absurd productivity claims and employees feeling pressured to use AI tools. This phenomenon is not only prevalent in the private sector but also affects government institutions, leading to a breakdown in effective decision-making. As a result, companies are no longer able to focus on anything important, unless they are one of the few organizations where AI happens to address their highest priorities.

SQLite's strict tables feature helps enforce rigid typing, preventing mistakes like putting text in number columns. Learn how to use strict tables and their benefits.

The EU has two proposed Chat Control laws, 1.0 and 2.0, which are moving through institutions in parallel, causing contradictory news. Chat Control 1.0 has been revived, while Chat Control 2.0 remains unagreed.

LG monitors have been found to silently install software through Windows Update without user consent, raising concerns about privacy and control over software installations.

A rare glimpse into Amazon's shadow bribery market has been revealed, with sellers being offered access to company employees for a price. The market, which operates on messaging apps like WeChat and Telegram, has been a persistent scourge of Amazon's online store.

Kaiser Permanente nurses claim AI-powered workplace surveillance is compromising patient care and their ability to provide empathetic support. As contract negotiations begin, nurses are pushing back against the use of AI in call centers, citing concerns over automated systems assessing their empathy and tone of voice.

PlayStation's European terms of service allow the company to delete accounts and purchased digital games after 3 years of inactivity, sparking concerns over digital ownership.

Apple has sent legal letters to dozens of OpenAI employees, directing them to preserve documents and communications, as the company pursues its trade secret lawsuit against the AI company.

Sony is deleting over 500 purchased movies from PlayStation Store accounts, highlighting the issue of digital ownership and licensing agreements.

OnePlus, a consumer electronics brand, will cease operations in the US and Europe as part of a larger restructuring under its parent company Oppo. Existing devices will continue to receive software updates and security patches, but no new products will be launched in these regions.

Payments company Stripe and private equity firm Advent International have made a joint offer to acquire PayPal Holdings Inc for $60.50 per share, valuing the company at over $53 billion. The offer is backed by $50 billion in committed financing from banks and represents a 28% premium to PayPal's closing share price on Tuesday.

A developer successfully ran Google's Gemma 4 26B model on a 13-year-old Xeon with no GPU, achieving 5 tokens per second decode speed. The journey to get there reveals the state of local LLM inference on non-standard hardware.

Thinking Machines Lab releases its first open-weights model, Inkling, with 975B parameters and multimodal input capabilities. The model is available for fine-tuning and integration into third-party products.

SpaceX's bond value has dropped to 10% less than its issue price, sparking concerns about the company's financial health and potential junk bond status. The bond market is signaling that SpaceX is a higher risk investment, with an average credit spread of 1.62 percentage points over Treasuries, similar to BB-rated bonds [1].

The AI investment boom is driving a shift in financing from cash flows to debt, with private credit playing a key role. This shift raises concerns about financial stability and the potential for systemic spillovers.

Dependabot now waits three days before opening a version update pull request, reducing the risk of supply chain attacks. This cooldown period is configurable and only applies to version updates, not security updates.

The era of free open-source software is coming to an end. Maintainers are burning out, and companies are exploiting their labor without contributing back. It's time to rethink the economics of open-source software.

A recent update to OpenAI's Codex CLI has introduced end-to-end encryption for inter-agent communications, making it difficult for users to inspect and debug their AI workflows.

New York becomes the first US state to impose a data center moratorium, citing concerns over power costs, water supplies, and local communities. The one-year moratorium will apply to data centers using 50 megawatts or more of power.

Scientists from Japan have developed a new method to recover up to 90% of lithium from used EV batteries, a breakthrough that could change how EV batteries are made and reused.

AWS will stop accepting new Mechanical Turk customers on July 30 2026, signaling the end of a once‑crucial crowd‑labelling platform for AI pipelines.

LARP lets founders record revenue without moving cash, but the legal and financial fallout may outweigh the headline‑grabbing ARR.

Since Chrome 148, the host libm backs Math.tanh, allowing anti‑bot systems to identify the underlying OS from a single floating‑point probe.

A side‑by‑side measurement shows Claude Code injects ~33k tokens before a prompt, while OpenCode needs only ~7k—fueling higher costs, latency, and cache churn for production AI agents.

A CSO report shows Ireland’s server farms surged to 23% of total electricity use in 2025, prompting new grid rules and fresh cost‑risk calculations for operators.

A tiny 2 GiB test node exposed a Go context leak in kubelet that can balloon memory usage, but profiling and a fast‑tracked patch resolved it for production clusters.

Kimi K2.7 Code brings an open‑weight model to GitHub Copilot, but enterprises must follow a concrete rollout plan to control cost, security, and compliance risks.

PeerTube offers a federated, ad‑free video platform that cuts hosting costs and eliminates vendor lock‑in for organisations seeking control over their media streams.

The ‘father of the Internet’ Vint Cerf announces his retirement from Google, prompting a look at his technical legacy and the business implications for standards‑driven AI.

Embedding workflow metadata in the same Postgres database as application data lets you achieve exactly‑once semantics and atomic cross‑system updates without extra outbox services.

ZeroFS and Amazon S3 Files both turn object storage into a POSIX mount, but their storage layouts, visibility semantics, and cost structures differ sharply for enterprise workloads.

A new analysis of 2.6 billion QuickDraw sketches shows that visual embeddings capture cultural nuances that text‑only models miss, reshaping how enterprises should train multimodal AI.

A UPI transfer looks like a single tap, but it traverses seven separate organisations before you see the green tick. Here’s how each layer works and what it means for cost, risk, and enterprise strategy.

A single OOM kill can bring down an entire PostgreSQL instance. Enforcing strict memory overcommit turns catastrophic crashes into graceful allocation failures.

A founder’s grand market thesis collides with engineering reality, turning a promising oven startup into a case study in product debt and misaligned incentives.

The new DMARC np tag, meant to block mail from non‑existent subdomains, can break when DNSSEC uses compact denial of existence, leaving organizations exposed.

Websites are under a fresh wave of traffic from compromised home devices, forcing operators to balance security with user experience.

New York City has adopted a new rule banning companies from using deceptive subscriptions, targeting companies that make it difficult for customers to cancel services or hide extra fees. The rule, which starts on October 1, promises hefty fines and aggressive enforcement for violators.

Mesh LLM is a distributed AI computing platform that pools GPUs and memory across machines, exposing a single OpenAI-compatible API. This approach allows users to run large language models without relying on expensive data centers or metered APIs.

Jim Keller's startup Fab2 is building a factory to mass-produce small semiconductor fabs, aiming to disrupt traditional chip manufacturing. The company's approach uses electron-beam lithography, which has a lower throughput compared to EUV scanners, but offers faster prototyping and lower costs.

A new AI tutor has demonstrated a significant impact on student performance in a Dartmouth course, with effect sizes ranging from 0.71 to 1.30 standard deviations. The AI tutor, called Phosphor, uses large language models to grade formative assessments and provide feedback to students.

With 6 million installs and zero‑tracking architecture, Organic Maps offers a privacy‑first, fully offline GPS solution that cuts data costs and simplifies compliance for remote workforces.

Mark Zuckerberg acknowledged that Meta's AI agent development has not progressed as quickly as expected, citing a slower-than-anticipated trajectory over the last four months.

Neoclouds are one of the more hotly debated AI business models, with CoreWeave and Nebius being the two most widely recognized names. These companies have seen their sales, backlog, and share prices soar, differentiating themselves through quick access to the latest GPU compute and GPU utilization advantages that allow hyperscalers to rapidly add efficient compute capacity.

The emergence of GLM 5.2, a high-quality open-weights model, threatens to disrupt the AI industry's profit margins. With costs potentially 50% lower than frontier labs, businesses may soon face a margin collapse.

The European Parliament has passed the Chat Control regulation, allowing for the scanning of private messages for child sexual abuse material. The move has been met with criticism from civil rights activists and IT security researchers, who warn of mass surveillance and unacceptably high error rates of AI scans.

Apple has sued OpenAI, accusing the company of stealing trade secrets. The lawsuit alleges that OpenAI's chief hardware officer, Tang Tan, and a former Apple employee, Chang Liu, misappropriated Apple's confidential information.

The European Parliament has approved an urgent procedure to fast-track legislation that would revive the EU's expired private message scanning rules, setting up a decisive vote on July 9.

Apple's Mac mini and Mac Studio have become the machines of choice for running AI agents, according to Doug Brooks, Apple's senior product manager of Apple silicon. Brooks made the claim while discussing Apple's chip strategy in a newly published interview with The Deep View conducted just prior to WWDC 2026 in June.

The European Commission has found Meta's Facebook and Instagram in preliminary breach of the Digital Services Act (DSA) due to their 'addictive' design. The Commission suggests that the platforms' design features, such as infinite scroll and autoplay, may harm users' physical and mental well-being.

John Deere owners will soon have the right to repair their own equipment under a new FTC settlement.

Meta has developed a custom bridge chip to reuse old RAM in new servers, reducing electronic waste and cutting costs. The technology, called Vistara, decouples older memory from server memory channels, enabling its reuse in new machines alongside native memory.

The European Parliament has allowed the suspicionless mass scanning of private communications to pass, despite a majority of voting Members of the European Parliament opposing the regulation. Dr. Patrick Breyer warns of the consequences, stating that 'our children lose out' in this undemocratic process.

Telemetry shows GPT‑5.5 responses clustering at exactly 516, 1,034 and 1,552 reasoning tokens, a pattern linked to recent drops in complex‑task performance.

Sony will stop manufacturing physical discs for any new PlayStation releases starting January 2028, shifting fully to digital distribution.

A Member of the European Parliament was hacked with Pegasus spyware while serving on a committee investigating spyware abuse. The incident highlights the threat of mercenary spyware to democratic institutions.

As satellite broadband fills the gap left by fragile terrestrial networks, African IT leaders need a clear rollout plan that balances cost, security and scalability.

Wafer AI achieves 213 tokens/second single-stream and 2626 tokens/second aggregate per node on AMD MI355X, outperforming Blackwell at over 2x lower cost. Learn how they optimized speculative decoding and tuned MoE kernels to reach this milestone.

A novel approach to reducing Fable costs by converting code to images and leveraging OCR technology. This method achieves a 60% cost reduction without compromising model performance.

A step‑by‑step breakdown of Jamesob’s $46,000 rig that delivers 384 GB VRAM and sub‑microsecond GPU interconnect for state‑of‑the‑art LLM inference.

Alibaba will ban employees from using Claude Code in workplace environments due to alleged security risks involving embedded backdoors, a source says. The move comes amid growing concerns over the potential risks associated with AI-powered coding tools.

The Spanish government has blacklisted US data analytics giant Palantir Technologies from public and private state-controlled companies due to growing concerns over the potential misuse of classified national security information. This move is part of a broader effort to preserve national data sovereignty and reduce reliance on foreign defense software.

Germany's Infineon has opened a major chip plant as the EU seeks tech autonomy, with the new facility producing chips for intelligent power management used in electric vehicles and data centers.

A new trojan dubbed Android Developer Verifier (ADV) is being installed via Play Protect, giving Google root‑level control and potentially disabling non‑Google apps worldwide.

Developers can’t reliably block sensitive files from Codex’s agent, exposing credentials and inflating compute costs. A .codexignore proposal remains unanswered.

Optimistic Provide reduces IPFS content publishing latency from 15 seconds to under 1 second, improving performance by over one order of magnitude.

The KIDS Act would require age checks to get online, pushing online services to verify users' ages and restricting online speech.

The Department of Commerce has lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 AI models, allowing the company to restore access to these advanced tools. This move comes after the US government had previously restricted access to these models over national security concerns [1].

US consumers and small PC firms allege the DRAM trio colluded to curb supply and inflate prices by 700% since 2022, raising antitrust concerns for the memory market.

Rocket Lab has signed a definitive agreement to acquire Iridium Communications in a cash-and-stock transaction valued at approximately $8 billion, creating a fully vertically integrated space and communications platform. The acquisition combines Rocket Lab's launch and satellite manufacturing capabilities with Iridium's global L-band constellation, spectrum rights, and over 2.55 million active subscribers.

Brown University’s ECON 1170 exam revealed at least 50 students used AI to cheat, prompting urgent calls for a campus‑wide integrity debate.

WAL-RUS is a Rust-based implementation of Postgres backup and WAL archival tooling, designed for predictable memory-efficiency and WAL-G compatibility. Learn how to deploy and use WAL-RUS in your Postgres environment.

A step‑by‑step guide to configure two AMD Strix Halo nodes with Intel E810 RoCE v2, unified memory, and a patched RCCL library for distributed vLLM inference.

California's proposed 3D printer surveillance scheme threatens privacy and choice for users, and its effectiveness is questionable. The scheme mandates print-blocking algorithms that can be easily circumvented, and its implementation may harm law-abiding users.

Asian AI startups are launching models that promise Mythos-like capabilities without fear of an export ban, potentially threatening US AI labs' market share. Sakana AI's Fugu and 360 Security's Tulongfeng are two examples of these new models.

Machine‑learning engineers have built reinforcement‑learning and diffusion models that generate radio‑frequency chips in minutes, slashing design cycles and unlocking performance beyond human‑crafted templates.

A deep dive into the PinpinRAT supply‑chain compromise reveals how a fake VC interview scam slipped a multi‑layered RAT into a TypeScript repo and why the error points matter for enterprises.

Qualcomm is buying AI startup Modular in a $4 billion all-stock deal, aiming to strengthen its position in the data center market with AI-native software. The acquisition pits Qualcomm against Nvidia's CUDA platform.

The Trump administration is now vetting every company that wants access to OpenAI’s latest model, GPT‑5.6, turning a once‑open AI rollout into a case‑by‑case licensing regime.

OpenAI’s limited preview of GPT‑5.6 Sol promises deeper reasoning and stronger safeguards, but enterprises must weigh new pricing and phased access.

Anthropic says operators linked to Alibaba used thousands of fake accounts to harvest billions of Claude interactions, prompting U.S. export controls on its newest models.

An internal security breach exposed keystroke and screen‑capture data from Meta's Model Compatibility Initiative, prompting the company to pause the program and reassess its AI‑training strategy.

A former Google engineer’s open‑source Workspace CLI went viral, hit #1 on Hacker News, and led to his termination—raising questions about internal policies and brand use.

A new poll shows 60% of Americans find the word “AI” in advertising off‑putting, forcing marketers to rethink how they frame technology benefits.

Anthropic’s latest identity verification layer for Claude reshapes AI security and compliance, but it also brings hidden costs and integration hurdles for large organizations.