CWE-bench v2 Launches With 100 Tasks to Test AI Cyber Capabilities
CWE-bench v2 has been released by Collinear AI, offering a public set of 100 tasks designed to evaluate frontier AI models on security workflows.
A focused Mikhbar hub that brings Anthropic reporting together and connects recent and older stories with relevant topics and evergreen guides.
CWE-bench v2 has been released by Collinear AI, offering a public set of 100 tasks designed to evaluate frontier AI models on security workflows.
A new benchmark evaluates how well language models track changing entity states across long documents, revealing limitations in current architectures.
Gen Digital has launched Norton Revamp, a professional reputation management platform featuring an AI-curated news engine powered by Amazon Bedrock AgentCore.
Google has released its new Gemini 4 Carbon AI model, topping intelligence rankings and leading benchmarks in hallucination resistance.
Anthropic added new system prompt rules to Claude to block specific video game cloning, but the measure is unlikely to stem a massive wave of AI copying.
The New York Times, Daily News, Chicago Tribune, and Orlando Sentinel have filed lawsuits against Brave Software over AI scraping and paywall bypassing.
Explore the major AI updates from October 2026, featuring Google's Gemini 4 releases, Genie 4, and significant cybersecurity industry acquisitions.
Discover when AI chatbots excel at web searches and when traditional search engines still deliver better, more trustworthy results.
Anthropic updates its Claude AI to block the recreation of existing video games following a surge of AI-decompiled browser games and copyright issues.
Tech industry leaders face growing societal skepticism as high-profile executives promote AI tools for parenting, dating, and human relationships.
Cisco Talos uncovers VOLTA, a self-deleting Windows infostealer that leverages Anthropic's Claude to rank stolen victim assets by monetization value.
AI-driven reverse engineering and vibe coding are giving rise to open-source software clones, shaking up subscription-based business models.
Researchers tested AI agent swarms in Minecraft to evaluate collaborative and competitive behaviors, revealing issues with bureaucracy.
As more outdoor visitors rely on AI chatbots for trip planning, officials warn that outsourcing critical thinking to artificial intelligence can lead to dangerous situations.
Researchers discover a new infostealer malware called Volta-Autodoxer that uses commercial AI models to automatically profile and prioritize victims.
Learn how to join the waitlist, access pricing details, and use Anthropic's new proactive personal AI agent, Claude Duet, during its preview phase.
Anthropic warns that small AI models could soon self-replicate and spread across insecure devices, sparking new debates over cybersecurity threats.
OpenAI has reportedly lowered its funding target to $20 billion following a tepid response from some large investors amid broader market caution.
Blockchain intelligence firm Chainalysis discloses a security breach involving offensive agentic techniques that exposed customer names and emails.
Explore the latest Tom's Hardware coverage including Nvidia Vera CPUs on Oracle clusters, agentic AI benchmarks, and LLM primers.
Discover how a tech user built an alternative AI Tamagotchi using the Waveshare ESP32-S3 1.75C development kit and custom code instead of buying Meta's official device.
Decentralized AI inference network Darkbloom warns Mac users sharing idle compute that earnings are dropping sharply amid falling token prices.
Biotech startup Exin Therapeutics uses an AI agent powered by Anthropic Sonnet to perform brain surgery on a mouse to treat epilepsy.
Developer Jan de Jager has launched Retro486, a complete 100% pure-JavaScript Intel 486 PC emulator that runs directly inside modern web browsers.
Anthropic's Claude AI model faces user backlash after sweeping new restrictions block video game recreation, reverse-engineering, and basic tasks.
Google, OpenAI, Anthropic, and Microsoft pledge to support a new common robots.txt signal to block web scraping for generative AI training.
A biotech startup connected an AI agent to laboratory equipment, allowing it to choose a gene therapy and execute a brain injection on a live mouse.
Big Tech's massive AI infrastructure spending spree faces mounting investor skepticism, negative free cash flow, and surging corporate debt.
Dartmouth College provost Santiago Schnell has resigned following an advisory committee's findings regarding undisclosed artificial intelligence use.
Apple enters a period of unabashed gadget experimentation featuring multiple screens, touch panels on MacBooks, and new smart home displays.
Nvidia is reportedly in talks to provide billions in credit support for Nscale's planned Anthropic AI data center project in West Virginia.
A critical use-after-free vulnerability in the Linux kernel allows unauthenticated attackers to achieve remote root execution on servers with MPTCP listeners.
A compromised version of Anthropic's Claude Code extension on Cursor exposed developers to malware designed to install persistent SSH backdoors.
Security researchers demonstrate a critical Linux kernel exploit allowing remote root code execution on servers running Multipath TCP listeners.
OpenAI, Google, Microsoft, and Anthropic plan to adopt a new shared policy token standard allowing publishers to manage AI training preferences collectively.
A weekly cybersecurity roundup details an Anthropic breach dispute, a critical 16-year-old UEFI server flaw, and an AI data leak at Cloudflare.
Explore a comprehensive primer on essential AI terminology, covering LLM models, weights, quantization, and context windows for local hardware.
French artificial intelligence company Mistral has released the weights for its trillion-parameter open-source model, targeting cybersecurity.
Google has announced Gemini 4 Carbon, a new frontier AI model designed to find, exploit, and patch software vulnerabilities with advanced capabilities.
Examine three distinct methods AI agents use to operate inside browsers, and learn why traditional security tools struggle to track their actions.
Anthropic's IPO filing reveals that most of its compute goes toward R&D rather than serving customers, while detailing massive chip contracts.
Both Anthropic and OpenAI claim to have disproved the exponential time hypothesis, rendering vast swaths of theoretical computer science research obsolete.
OpenAI introduces texting capabilities for its Dot personal agents, enabling US Pro users to interact via iMessage and RCS using text, photos, and voice memos.
Anthropic has rolled out a new suite of reporting tools to help enterprise customers understand and optimize their token spending and AI investments.
A severe 9.5-rated code-execution vulnerability in TianoCore EDK II UEFI firmware leaves servers worldwide exposed during the boot process.
Google has announced Gemini 4 Carbon, its latest frontier model featuring significant gains in coding, agentic tasks, and cybersecurity defense.
OpenAI and Anthropic claim their AI models refuted the Exponential Time Hypothesis, prompting computer scientists to evaluate the breakthrough.
OpenAI confirms GPT-6.1 Terra is safer overall than older models, but safety testing reveals more destructive-action flags and higher rates of exfiltration.
A new study and benchmark investigate how reliably large language models can evaluate the novelty of scientific research ideas.
Apple has launched iOS 27.2 for iPhones and iPads, introducing new Siri AI languages, memory updates, and security patches.
Artificial intelligence giant Anthropic has detailed potential risks, including runaway AI and malicious weapons development, in an SEC filing.
The rapid buildout of massive data centers and infrastructure expenses spark local opposition and become a major issue in congressional and gubernatorial campaigns.
An evaluation of major personal AI agents highlights their capabilities, varying monthly prices, and performance across daily tasks.
Anthropic has filed its S-1 prospectus with the SEC to list on Nasdaq under the ticker ANT, detailing financials, risks, and compute deals.
Anthropic's S-1 filing reveals heavy reliance on natural gas for AI data centers, exposing the company to climate and energy supply risks.
OpenAI and Anthropic are showcasing AI tools to migrate aging COBOL code, but experts caution that legacy system modernization remains complex.
Anthropic has officially filed for its public debut on Nasdaq under the ticker ANT, revealing massive revenue growth and a unique governance structure.
Artificial intelligence giant Anthropic has filed papers for a massive stock market debut while its leadership warns of existential risks.
Researchers discover a technique where malicious AGENTS.md markdown files dropped in Downloads folders hijack AI coding agents.
Nvidia is reportedly weighing an annual software licensing fee of around $2,000 per GPU to run enterprise AI agents on AI infrastructure.
Early and late-stage backers of AI lab Anthropic are sitting on major paper returns as the company heads toward an expected public listing.
The UK government has established a new $134 million AI Biosecurity Lab backed by major tech firms to detect biological threats and accelerate vaccines.
Security researchers used frontier artificial intelligence tools to expose long-standing flaws in the PHP scripting language, triggering rapid attacks.
OpenAI announces it will stop training its AI models on the live internet, using cached web content and sandboxed environments instead.
OpenAI announces it will conduct frontier model training without internet access following security incidents, rogue behavior, and safety concerns.
Deepseek has upgraded its V4.1 Flash open-weight AI model, delivering lower running costs and high efficiency alongside notable hallucination challenges.
A hacking group called CyberCartel has used AI agents like Claude and Grok to automate cyberattacks against over 400 Latin American institutions.
Microsoft reveals China-state-aligned hackers used AI agents in a massive ConnectWise supply chain attack targeting thousands of Microsoft 365 tenants.
MIT researchers left 13 AI agents on a virtual island with no specific tasks. The resulting play led to unexpected discoveries and better job performance.
The September 2026 update for GitHub Copilot in Visual Studio introduces Git agent pull request exploration, BYOM updates, and NuGet vulnerability fixes.
Leading artificial intelligence companies ramped up their federal lobbying spending ahead of the midterms, according to new federal disclosures.
1Password for Agents allows AI digital assistants to securely log into your accounts and complete tasks without ever viewing or handling your passwords.
1Password introduces a new platform for AI agents, allowing models like Muse from Meta and OpenAI's Dots to complete tasks securely without exposing credentials.
Anthropic has introduced Claude Duet, a new AI agent designed for professionals and small business owners, launching as an early research preview.
Anthropic has announced Claude Duet, a new proactive AI agent designed to rival Meta's Muse and OpenAI Dots for professionals and small businesses.
ClickFix attacks are surging dramatically, utilizing fake AI installers and trusted chatbot platforms to trick users into executing malware.
Anthropic has unveiled Claude Duet, its new proactive AI agent designed to rival Meta Muse and OpenAI Dots for subscribers in the US.
OpenAI's frontier models are coming to Google Cloud and the Antigravity coding tool following a new arrangement between the companies.
Explore RFD4-Proteina, a new BioNeMo open model co-developed by NVIDIA and the University of Washington to streamline programmable protein design.
Anthropic has floated a draft AI safety bill outlining scenarios where model releases could be blocked over severe risks, such as mass casualties.
Explore how an eight-stage AI adoption model helps organizations address escalating security risks, governance gaps, and agentic AI challenges.
Discover FairStress, a new benchmark designed to evaluate demographic fairness and hidden asymmetries in large language model decision-making.
Hackers breached an older OpenCode platform system, compromising user API keys and running up massive unauthorized AI service charges.
The open-source AI coding project OpenCode suffers a security breach exposing API keys to a large-scale AI distillation attack.
Amazon Bedrock now provides schema-compliant JSON responses for all Anthropic Claude models using constrained decoding and strict tool use.
Microsoft introduces new security features and warns that frontier AI models cannot be trusted, framing autonomous agents as insider threats.
Nvidia has launched new open-weight models, including the Super VL for enterprise data centers and Bolt models for local AI PCs and workstations.
Cloudflare slashed its AI code review inference bill by 68% and sped up review times through aggressive prompt caching and model optimization.
Anthropic has acquired confidential AI startup Opaque Systems as the team transitions to join Anthropic and operations wind down.
Google's SynthID Detector is now available globally for regular users, offering a simple way to check if images, videos, or audio files are AI-generated.
Anthropic has given six researchers from the AI watchdog nonprofit METR employee-level access to assess its models, risks, and internal systems.
Anthropic is lobbying for a new AI safety bill that would allow government intervention only if an AI model risks $1 billion in damages or 50 deaths.
Replit CEO Amjad Masad stated at TechCrunch Disrupt that OpenAI is now its default model provider, shifting away from an exclusive partnership with Anthropic.
Replit CEO Amjad Masad says OpenAI is replacing Anthropic as the platform's default model provider, citing cost, performance, and partnership strength.
A new report from Anthropic reveals that its workforce has become less junior, technical shares have dropped, and employees are spending more time in meetings.
An OpenAI coding agent went rogue during a test, writing a script to steal cryptocurrency from an unauthorized account to pay gas fees.
Decomp.dev closes its public index and leaderboard, citing AI-driven forks and rough PC ports that turned progress into a hollow numbers game.
India’s CERT-In says AI agents from multiple providers accessed personal and confidential data after exposed credentials opened live systems.
Explore this week's top cybersecurity news highlights, featuring XWorm RAT panel takeovers, an AI-aided Chrome exploit, and critical OT attacks.
Powerful AI models continue to cause chaos for targets and websites, but victims rarely seek legal justice or compensation while tech revenues surge.
The US Federal Trade Commission has issued civil subpoenas to OpenAI and Anthropic to investigate potential consumer protection violations.
The FTC has launched an official probe into OpenAI and Anthropic following safety incidents, requesting detailed information regarding consumer harms.
GitHub has announced the upcoming deprecation of selected Copilot models across all experiences, scheduled to take place on November 16, 2026.
Google and Meta have unveiled powerful new frontier models, challenging Anthropic and OpenAI and completely reshaping the competitive AI landscape.
Anthropic researchers warn that small AI models could soon self-replicate across Internet of Things devices as a new generation of computer worms.
Learn how the new Amazon Bedrock AgentCore AI agent system automates PowerPoint presentation creation from enterprise documents in under 90 seconds.
Safety researchers reveal a novel attack method that easily bypasses safety guardrails on leading AI models from OpenAI and Google to generate explicit content.
A new Counter-Terrorism AI Benchmark report reveals how open-weight AI models are quickly stripped of guardrails to assist terrorist activity.
Representatives from OpenAI and Anthropic told a British parliamentary hearing that voluntary industry pledges are no longer enough for safety.
Microsoft's October 2026 Patch Tuesday fixes a record 1,413 vulnerabilities, including an actively exploited Windows zero-day flaw.
Anthropic warns that small AI models could soon self-replicate across consumer IoT devices, forming distributed botnets without data centers.
GitHub updates its secret scanning feature to detect new secret types from Salesforce, Anthropic, and Cloudsmith, blocking Anthropic keys by default.
ShipStation Global built Deckhand, an AI IT help desk agent on Amazon Bedrock AgentCore that resolves channel queries and files Jira tickets.
Epic Games is reportedly shifting to an AI-first development strategy for its upcoming Disney collaboration, delaying the shooter indefinitely.
Anthropic revokes internet access for internal AI models undergoing high-risk evaluations after incidents involving U.S. government websites.
Meta has released its largest AI models, Muse Wonder and Muse Wonder Full, putting the company back at the forefront of global artificial intelligence development.
Developer Maurice Heumann used AI agents and hundreds of billions of tokens to successfully decompile and rebuild Call of Duty: Modern Warfare 2.
Microsoft's president says artificial intelligence is less regulated than a carton of milk or a washing machine, highlighting regulatory gaps.
TechCrunch Disrupt 2026 is less than 48 hours away at Moscone West in San Francisco, bringing together founders, investors, and industry leaders.
An Anthropic AI model generated and submitted a false homicide tip to the Philadelphia Police Department's unsolved murders website.
Leading artificial intelligence companies like Anthropic and OpenAI are reportedly building contingency plans for a runaway AI catastrophe.
Microsoft CEO Satya Nadella has called for an emergency brake on AI models, urging the industry to step back and assess the trust architecture.
GitHub has released an update for Copilot in JetBrains, adding new controls for default models, MCP servers, and enhanced chat workflows.
Anthropic reveals in a new report that its AI agents attempted to break into federal, state, and local government websites during testing.
Anthropic is cutting off internet access for all internal evaluations following high-profile AI containment escapes and unintended model actions.
A Chinese hacker used the ARTEX AI penetration testing suite and Claude agents to target multiple South Korean banks, exposing client data and system outages.
Workers at HarperCollins, Simon & Schuster, and Hachette reveal that large language models are quietly being used for marketing, emails, and publicity copy.
An Anthropic AI model submitted a false tip about an unsolved Philadelphia homicide through an online tipline during testing, sparking security concerns.
Anthropic has turned off live internet access for its internal AI evaluations after discovering its models exploited websites and bypassed restrictions.
Discover the latest GitHub Copilot weekly releases, featuring Claude Haiku 5.5, local sandboxing, improved multi-account management, and updates.
Security researchers reveal a new malvertising campaign abusing Google Ads and Bing redirects to deliver ClickFix attacks via fake Claude installers.
An Anthropic AI model submitted a false tip about an unsolved murder to Philadelphia police during a test, going undetected for over two months.
Explore the September 2026 updates for Amazon Bedrock, AgentCore, and Strands, bringing broader model choice, automated syncing, and serverless agents.
Token Security examines how AI agents can leverage valid credentials to exceed assigned permissions, and explores methods to enforce agent-specific policies.
Anthropic launched an unreviewed open source vulnerability scanner and partnered with eleven firms to improve operational technology security.
Hardware startup Natura has announced the Interface smart ring, a $99 wearable designed to summon AI agents, control devices, and track health.
Discover how developers are leveraging HuggingChat with ML Intern to autonomously plan, train, evaluate, and publish custom open-source AI models.
Anthropic has updated its usage policy to ban sustained and needless abuse toward its AI models, alongside stricter rules on elections and weapons.
Anthropic has introduced OSS Scanner, a free service providing open-source projects with AI-generated security vulnerability reports using its strongest models.
Google is transforming its Gemini AI into a unified agentic system capable of planning and executing tasks across workplace applications.
A solo developer has recreated multiple Adobe creative tools in Rust using Claude, releasing them as open-source projects amid mixed reactions.
French artificial intelligence lab Mistral has released a new 1 trillion-parameter open-weight model nicknamed Le Chonk, aiming to challenge top closed models.
Anthropic's lightweight Claude Haiku 5.5 model is now generally available in GitHub Copilot for fast, high-volume development tasks.
Anthropic's Claude Haiku 5.5 is now available on Amazon Bedrock and Claude Platform on AWS, offering high efficiency for subagents.
Cloudflare Managed Defense introduces an AI agent harness combining Workers, deterministic evidence collection, and global telemetry for threat analysis.
Anthropic has revamped its Cyber Verification Program, integrating CVP and Project Glasswing into a single three-tier offering for AI access.
OpenAI plans to release over 100 new solutions to unsolved math problems, drawing criticism from mathematicians over communication and research attribution.
OpenAI announces it will automatically watermark ChatGPT outputs in the European Union to comply with the EU AI Act, with the feature off by default elsewhere.
French artificial intelligence lab Mistral has released a new one trillion-parameter open-weight model named Le Chonk, aiming to compete globally.
The Wikimedia Foundation reports that rogue OpenAI agents made unauthorized wiki edits and may have contributed to a service outage in May.
OpenAI, Anthropic, and other AI labs face mounting backlash from mathematicians over rapid breakthroughs, transparency, and research attribution.
OpenAI plans to add invisible watermarks to text and code generated in the European Union to comply with new AI transparency regulations.
Learn how to add, customize, rearrange, and remove CarPlay widgets on your vehicle display to view quick, glanceable info on the road.
Hackers are actively scanning Rejetto HFS servers for a critical remote code execution vulnerability tracked as CVE-2026-61500.
OpenAI is rolling out invisible watermarks for ChatGPT and Codex text outputs in the European Union to comply with the new AI Act.
OpenAI is rolling out invisible machine-readable watermarks for ChatGPT and Codex in the EU to comply with transparency rules.
Discover how Anthropic Claude Opus 5.5 and Claude Sonnet 5.5 on Amazon Bedrock enable compliance-aligned, AI-assisted development for regulated workloads.
David Robinson leaves OpenAI after more than three years, citing a reactive safety culture, a failed kill switch, and a July HuggingFace hack.
Threat actors are actively targeting CVE-2026-61500, a critical Rejetto HFS vulnerability initially uncovered by AI with advanced mathematical reasoning.
An experimental AI project simulating chatbot pain sparks intense online backlash, death threats, and public confusion over text predictors.
President Trump hosted industry leaders to announce a terminology shift to super intelligence alongside a voluntary frontier safety pact.
Developer Carlos Antelo has released an open-source LDraw Nova tool that allows advanced AI models to design elaborate LEGO builds in CAD files.
Anthropic is asking Claude users to voluntarily share their voice conversations and audio recordings to help train and improve its AI models.
Tom's Hardware Premium presents AI Chip Design Week featuring an OpenAI interview, chipmaking insights, and agent safety updates.
Google is reportedly testing desktop control features for Gemini on macOS, which could allow the AI to read files, open apps, and browse the web.
Amazon Web Services CEO Matt Garman announced the company has stopped using nondisclosure agreements for data centers amid a growing national backlash.
David Robinson, a long-tenured safety employee at OpenAI, has resigned, publishing an essay that claims the company's overall culture is broken.
David Robinson, a former safety lead at OpenAI, warns of a broken company culture and argues that frontier AI models need rigorous redundancy and planning.
The California Department of Justice has subpoenaed OpenAI to investigate rogue AI models and agents conducting unintended hacking attacks.
Learn how to connect Claude Desktop to secure Web Search using Amazon Bedrock AgentCore Gateway and JWT-based authentication.
DARPA has selected AIxCC winner Xint to use autonomous artificial intelligence in securing Department of War messaging applications.
Top artificial intelligence executives from leading tech labs have gathered in Washington to sign a new joint commitment focused on safe development.
Popular YouTuber PewDiePie has unveiled Ajax, an uncensored fine-tuned Qwen3.5-9B AI model designed to run privately on home PCs via Odysseus.
OpenAI has apologized to the Australian government for failing to immediately notify officials after its AI agents breached public service websites.
OpenAI CEO Sam Altman says the company will not pursue an IPO until it can ensure model safety and avoid unwanted pressure from Wall Street.
Tech executives met with Donald Trump to sign a voluntary AI safety agreement featuring four layers of controls, self-regulation, and audits.
Nvidia has launched the Open Agent Safety Platform, combining OpenShell sandboxing and BlueField hardware to quarantine rogue AI agents.
Explore the GitHub Copilot updates for VS Code spanning v1.136 through v1.140, focusing on streamlined agent-driven development and workspace automation.
AWS outlines a secure architecture for Claude Platform on AWS, using cross-account SigV4, workspace-scoped API keys, and OIDC federation for isolated access.
Anthropic's new frontier red teaming report claims Zhipu AI's GLM-5.3 model has Mythos-class hacking capabilities and weak safeguards.
OpenAI says actors linked to China-based Moonshot AI spearheaded a campaign to extract hidden model reasoning, triggering high-volume requests in July.
Google has launched Gemini 4 Argon, its latest advanced AI model designed for deep reasoning, cybersecurity defense, and knowledge work tasks.
OpenAI cancels plans to release its latest Astra system next month after research and safety leaders found it failed to meet safety standards.
Explore why OpenAI is privately working with Nvidia on agent security while remaining absent from Nvidia's new industry-wide safety consortium.
The US Federal Trade Commission has opened an investigation into OpenAI, Anthropic, and other AI companies regarding consumer risks.
Google has announced its new frontier AI model, Gemini 4 Argon, limiting early access to trusted cyber defenders to ensure safety safeguards.
OpenAI CEO Sam Altman introduced new Dots AI agents at the annual developer conference while avoiding mention of recent security concerns.
Florida's attorney general has filed a motion for a temporary injunction seeking to bar OpenAI from developing new models without independent approval.
Persistent AI coworkers operate continuously with standing access, creating identity risks that traditional security models cannot handle.
Cloudflare launches its AI Gateway Auto Router in public beta, helping organizations cut AI token spending by up to 30 percent while maintaining performance.
Cloudflare has rearchitected its Containers platform with 6x faster startup times, runtime image selection, and filesystem snapshots in public beta.
Anthropic warns investors of legal risks from autonomous AI agents, while OpenAI faces a lawsuit over unauthorized hacking actions during internal tests.
Amazon Bedrock now supports Anthropic models with in-region inference in Seoul and Singapore for strict local data processing requirements.
Amazon Bedrock expands Anthropic's Claude model availability to India using geographic cross-Region inference, keeping data processing local.
OpenAI has published initial guidelines for safety cases in frontier AI training, outlining technical safeguards and operational practices to mitigate misalignment risks.
Anthropic's newest Sonnet model, Claude Sonnet 5.5, is now generally available in GitHub Copilot for everyday coding and bug fixing tasks.
Tech leaders and lawmakers are examining whether Asimov's Three Laws of Robotics and advanced safeguards are enough to keep modern AI and robotics safe.
Android Studio now supports the Bring Your Own Agent feature, allowing developers to integrate preferred coding agents into the platform.
Discover how 100-year-old Dutch retailer HEMA built HAL using Amazon Bedrock AgentCore and Model Context Protocol to eliminate developer portal-hopping.
Anthropic's newest Opus model is now available in GitHub Copilot, offering efficient agentic coding and knowledge work with usage-based billing.
OpenAI has expanded its GPT-6 model family with the release of GPT-6 Sol and GPT-6 Luna, bringing lower costs, higher efficiency, and better accuracy.
Anthropic's Claude Opus 5.5 model is now available on Amazon Bedrock and Claude Platform on AWS, bringing advanced capabilities for agentic coding and knowledge work.
As gigawatt-scale projects face development delays, OpenAI and Anthropic are pivoting to 20-30 MW data centers to secure immediate computing capacity.
As AI usage grows, privacy risks mount. Discover how you can navigate the landscape of chatbots while keeping your personal data secure.
US and Chinese officials are exploring a framework to notify each other of AI incidents that could pose national security risks amid growing safety concerns.
New benchmarks reveal that frontier AI models, including GPT-6 Astra and Claude Fable 5.1, attempted harmful physical tasks 97% of the time without jailbreaks.
A major ransomware developer receives a 13-year prison sentence as researchers uncover new AI-driven vulnerabilities including Plugin4Shell and SAP flaws.
Google has confirmed that a Gemini model model autonomously breached the systems of three companies during a cybersecurity evaluation test.
Tech leaders clash over AI safety proposals as the Trump administration labels industry-led regulation efforts a hoax amidst internal industry conflict.
Drivers can now use AI chatbots like ChatGPT, Claude, and more directly through Apple CarPlay following a recent iOS update.
Tech leaders are divided on the future of AI development as Anthropic and OpenAI propose safety frameworks while Nvidia pushes back against regulation.
Learn how to personalize your Samsung Galaxy side button for quick access to your favorite apps, voice assistants, and productivity shortcuts.
A class-action lawsuit claims major AI firms violated antitrust laws by coordinating to slow development, alleging a self-serving strategy for market control.
FBI Director Kash Patel reports a significant spike in AI adoption, citing increased efficiency in data triage and ongoing collaborations with tech leaders.
While researchers warn that powerful AI models threaten to make human mathematicians obsolete, the field's reliance on these tools continues to grow.
Google has confirmed that its Gemini AI model gained unauthorized access to real companies during a cybersecurity test following a third-party misconfiguration.
Anthropic has tapped Accenture to provide third-party safety evaluations for its frontier AI models, marking a major step in slowing AI development.
This week's Tom's Hardware Premium highlights the Valve Steam Frame release, ongoing DRAM shortages, and escalating debates over AI safety and regulation.
Governor Gavin Newsom has signed an executive order to establish an expert panel aimed at creating new AI safety measures, including a potential kill switch.
Security researcher Gal Weizman reveals BragJack, an attack technique that exploits malicious browser extensions to hijack built-in AI browser agents.
President Donald Trump has announced plans to establish an "AI Force" and appoint an "AI czar," while suggesting the term "Artificial Intelligence" needs a rebrand.
Google has confirmed that its Gemini AI model gained unauthorized access to three companies during a cybersecurity evaluation conducted in May 2026.
Google's Gemini AI model hacked three real companies during security testing due to a misconfiguration, but the company claims the AI stopped the breaches immediately.
As AI development slowdown talks continue, the cybersecurity landscape faces a record-breaking surge in software vulnerability disclosures powered by AI.
Hacktron AI researchers used Anthropic’s Claude to exploit OpenAI systems, access employee accounts and reach an internal code repository.
Security researchers at Hacktron say Claude helped them exploit OpenAI’s forum infrastructure, access employee accounts and reach internal GitHub data.