All News

2219 articles • Page 41 of 148

AI News & TrendsMay 23, 2026

GitHub breach exposes 3,800 repos via malicious VS Code extension

A GitHub breach may have exposed about 3,800 internal projects after an employee installed a malicious VS Code extension linked to TeamPCP. The attack appears to have only stolen company code, and GitHub quickly rotated secrets and isolated the infected computer. Investigators have not yet shared the full attack method, but experts suggest that reused tokens and extension malware might be involved. Security teams are advising better control of developer tools and careful monitoring. It remains uncertain if attackers tried to stay inside GitHub systems or if more attacks may follow.

AI News & TrendsMay 22, 2026

Jury clears OpenAI in Musk lawsuit, removes IPO hurdle

A jury found against Elon Musk's claims toward OpenAI, which may remove a big legal barrier to the company's possible public listing. However, OpenAI still faces several other lawsuits and regulatory questions that could affect its plans. The company appears to be preparing for an IPO later this year, but timing might change if legal or regulatory issues grow. Ongoing cases and OpenAI's unique structure could still influence whether or when it goes public. Further developments may shape what investors think before any official stock market filing.

AI News & TrendsMay 22, 2026

Shai-Hulud Worm Forges Sigstore Provenance, Infects npm and PyPI

The Shai-Hulud worm quickly spread through npm and PyPI packages, forging Sigstore provenance to look like trusted builds. Investigators say the attack mixed old account takeover tricks with new ways to hijack CI pipelines, and did not need stolen credentials at first. The worm's payloads could steal credentials, set up persistence in code editor configs, and sometimes wipe disks if the computer matched certain locations. Experts suggest this attack may have both money and geopolitical motives. It is still unclear how common forged provenance will become, but early signs show attackers may prefer tricks that fool trust checks over brute force methods.

AI News & TrendsMay 22, 2026

OpenAI's GPT-5.2 integrates verifiable reasoning; Anthropic nears $30B revenue

OpenAI's GPT-5.2 now combines long-term reasoning with more careful checking of its answers, and can help with formal proof attempts, though expert review is still needed. Around the same time, Anthropic reported deals for a large amount of computing power and said its yearly revenue rate may be close to $30 billion. Benchmarks for AIs now focus on answers that can be checked for correctness, but only about half of recent AI-made proofs earned high confidence from experts. There may be risks in assuming AIs are always right, and companies might need new ways to check their work and plan for changing costs. The future of AI seems to depend on how proof, computing power, and company organization work together, not just on any one of these factors alone.

Business & Ethical AIMay 22, 2026

SAP Explains Business AI Decisions With New Audit Features

SAP has introduced new audit features to make its Business AI more explainable and trustworthy. The company warns that AI, even when it sounds confident, can make mistakes without enough context. The latest updates may allow users to see the reasons behind AI decisions, such as what information was used and how confident the system was. Regulators and leaders appear to be demanding more proof and transparency, with firms expected to keep logs and provide explanations for key decisions. These changes suggest that making AI decisions traceable and understandable is now an important part of business operations and compliance.

AI News & TrendsMay 22, 2026

White House Proposes 90-Day AI Model Review Before Public Release

The White House has proposed a 90-day review period for advanced AI systems before they are released to the public. This review may be voluntary at first, but could become a standard if major AI companies take part. The plan suggests government agencies would check these AI models for safety and security risks, focusing on threats like cyber or chemical misuse. Some details, such as how the review would work and if it will always be voluntary, remain unclear. Companies may need to adjust their schedules and prepare for possible delays if the order is signed soon.

Business & Ethical AIMay 22, 2026

NanoClaw pivots to enterprise AI, secures agents with Rust gateway

NanoClaw is shifting to focus on enterprise AI by making a secure, open-source tool that helps companies safely use AI agents at work. The system uses containers and a Rust gateway to keep agents isolated and only allows actions after checking company rules, with some actions needing human approval. Reports suggest NanoClaw may fit teams in finance, healthcare, and legal fields that need strong security and clear tracking of agent actions. The platform has gathered about 250,000 downloads and 29,000 GitHub stars, which may show growing interest, though it is not clear how many use it in production. Analysts suggest its strong security approach sets it apart from bigger cloud competitors, but it might face more competition soon.

AI News & TrendsMay 22, 2026

GitHub employee installs malicious VS Code extension, exposing 3,800 repos

A GitHub employee may have exposed about 3,800 internal projects after installing a malicious Visual Studio Code extension linked to TeamPCP. The attacker appears to have cloned these private repositories, but GitHub says there is no evidence that customer data was leaked. Security experts suggest the attack is part of a larger Shai-Hulud campaign that targets software supply chains. Investigations are ongoing, and it is not clear what specific data was taken from the repositories. This case suggests companies should be careful with third-party developer tools, as they may create security risks.

AI News & TrendsMay 22, 2026

Russia embeds AI in malware, uses 'AI poisoning' against Ukraine

Ukrainian officials say that Russia is increasingly using artificial intelligence (AI) in its malware and is trying 'AI poisoning' tricks, based on reports reviewed by security experts. This may mean that Russian hackers are making their attacks faster and harder to detect by changing how their malware and commands work. Some experts believe that Russia might also try to trick the AI defenses used by Ukraine by feeding them bad data, which can cause security systems to miss real threats. There is still little public technical detail about these attacks, but many worry that AI-powered cyberattacks could get worse and harder to defend against.

AI News & TrendsMay 22, 2026

OpenAI nets nearly $6 billion in Q1 2026, still unprofitable

OpenAI made almost $6 billion in revenue in the first quarter of 2026 but still lost money, with a very negative profit margin. Anthropic, a rival company, made about $4.7 billion that quarter but might reach a profit as soon as the next quarter, while OpenAI remains unprofitable. Even though OpenAI has more users and higher revenue, its costs are very high, and forecasts suggest it could lose $14 billion in 2026. There may be an OpenAI IPO soon, but plans are not certain yet. Both companies' future profits remain unclear and depend on how well they can manage costs and keep growing. forecasts suggest it could lose $14 billion in 2026.

Institutional Intelligence & Tribal KnowledgeMay 22, 2026

Amca Closes $300M Series B for Industrial Scale, Valued at $1B+

Amca recently closed a $300 million Series B funding round at a reported valuation above $1 billion. This round may signal that investors are willing to fund growth when manufacturing processes and product quality appear stable. Industrial hardware startups often face unique challenges, such as long certification cycles and higher costs from tariffs, which may require careful planning and capital use. Investors seem to want proof of consistent production and clear financials, and many companies structure funding in phases tied to specific goals. Experts suggest keeping extra capital on hand for surprises and using a mix of funding sources to limit ownership loss.

AI News & TrendsMay 22, 2026

Google Expands Demand Gen with AI-Powered YouTube Creator Tools

Google has expanded Demand Gen with new AI-powered tools for YouTube creators, which may help turn videos into direct response ads. The update includes AI-assisted setup, new ways to measure branded searches, and tools to test different ad versions. Features like Shoppable CTV and Travel Feeds suggest advertisers can now reach users on more devices and show live product details. Some early analysis from Google suggests brands using several best practices saw about 40 percent more conversions, but sample sizes were not shared. Regular updates mean advertisers may need to check settings often as new features appear.

AI News & TrendsMay 22, 2026

X unveils Creator Connect, expands AI-powered brand partnerships

X has introduced Creator Connect, a new feature that uses AI to help brands find and work with creators who match their campaign goals. This move may be driven by a possible rebound in advertising revenue and increased interest in performance data. Early reports suggest creator earnings are mainly concentrated among top accounts, and most creators make less than $400 a year from X. The platform appears to encourage creators to maintain premium subscriptions, get high organic reach, and make various types of content. Some experts believe AI matching could help brands find effective niche creators, but there are concerns it might make creator voices too similar or focus too much on quick results.

AI News & TrendsMay 22, 2026

Juicebox launches AI agents to source talent, promises 5x recruiter efficiency

Juicebox has launched AI agents that may help recruiters by automatically searching large databases and running outreach to find job candidates. The company says these agents can keep candidate lists fresh and may increase recruiter efficiency by up to five times, though exact numbers are not public. The system appears to work by scanning profiles, sending emails, and learning from recruiter feedback. Analysts suggest these tools could save time but also warn about possible fairness and privacy concerns. Juicebox's system is described as a helper for recruiters, not a replacement, and some uncertainty remains about its broader impact and oversight.

AI News & TrendsMay 22, 2026

Apple unveils new AI features for VoiceOver, Magnifier, Vision Pro

Apple announced new AI-powered accessibility features coming later this year for devices like iPhone, iPad, Mac, and Vision Pro. VoiceOver may give better scene descriptions, Magnifier appears to answer questions about nearby objects, and Voice Control might let users speak more naturally. The updates only work on newer Apple devices that support Apple Intelligence, so some older devices may not get them. Apple also says devices will soon create subtitles for any video in English, and Vision Pro may help control certain wheelchairs with eye-tracking. The release timing is uncertain, and not all details are final.