Content Fans

All News

2214 articles • Page 27 of 148

OpenAI uses Codex to migrate 600 petabytes in two months
AI News & Trends

OpenAI uses Codex to migrate 600 petabytes in two months

OpenAI used Codex, a large language model, to help move about 600 petabytes of data and rebuild 10,000 workflows in just two months. Codex generated scripts and checked data as it moved between cloud providers, which may have cut development time by about half. Engineers added approval steps and safety checks at risky points, suggesting that careful human review is still important. The results show Codex-style automation might soon be common in big data projects, but human oversight seems necessary for safety. Error rates were very low, and most problems were fixed quickly, which appears to match or beat usual manual methods.

Microsoft Unveils Four IQ Services for Enterprise AI Agents at Build 2026
AI News & Trends

Microsoft Unveils Four IQ Services for Enterprise AI Agents at Build 2026

Microsoft has introduced four new IQ services - Web IQ, Work IQ, Foundry IQ, and Fabric IQ - that may help enterprise AI agents work better by providing key information in separate layers. These services aim to make agent setup faster, safer, and simpler by handling different types of data such as live web content, organizational connections, business knowledge, and company metrics. Early feedback suggests these tools might speed up development, improve information quality for agents, and give IT teams better control. However, some details about how these services work and how easy they will be to use with other platforms are still unclear. Experts say that future adoption will likely depend on how well the new APIs perform and whether outside platforms can use them easily.

Enterprises Formalize Shadow AI, Cut Hours, Shorten Cycles
Business & Ethical AI

Enterprises Formalize Shadow AI, Cut Hours, Shorten Cycles

Generative AI tools are being used in many workplaces before official rules are set, which may boost productivity but can also increase risks if not managed. Some evidence suggests that when companies formally add approved AI tools and train their teams, they can save time and shorten work cycles. However, these benefits might not be fully realized unless leaders change roles and track how time saved is used. There are also signs that sharing AI successes openly helps build trust and reduces employee resistance. Overall, the text suggests organizations should guide and measure AI use to balance innovation with security and compliance.

New Data Project Aims to Standardize AI Token Usage Metrics
AI News & Trends

New Data Project Aims to Standardize AI Token Usage Metrics

A new Data Project aims to help companies measure how many AI tokens are creating real economic value. The project suggests collecting anonymized data on token usage and product outcomes across different firms. This may allow leaders to compare their AI investments and spot trends in adoption and productivity. Early benchmark data suggests AI referrals might convert at higher rates than traditional search, but direct links between tokens and value are still unclear. The project hopes to create shared metrics that let companies track and compare their AI impact more accurately.

OpenAI Codex tops 5M weekly users, expands into knowledge work
AI News & Trends

OpenAI Codex tops 5M weekly users, expands into knowledge work

OpenAI says Codex now has more than 5 million weekly users, a big increase since February. About 20 percent of users may be knowledge workers, not just developers. Codex is being used for more than just coding, as it is now part of many business tools and workflows inside ChatGPT. Studies suggest it might boost productivity for both developers and business users. It appears that more companies are starting to use Codex regularly, but it is not certain if the high growth will continue.

AI Hallucinations Spur Slopsquatting Threat for Software Supply Chains
AI News & Trends

AI Hallucinations Spur Slopsquatting Threat for Software Supply Chains

Researchers say AI tools sometimes suggest fake software package names that sound real, and criminals may register these names to spread malware through automated build systems. This new risk, called slopsquatting, appears to be growing in 2025, but there have not yet been large confirmed attacks. Some experts warn that the threat is credible because automated systems might trust these fake packages without review. Security advice suggests using stricter controls like lockfiles, human approval, and monitoring for suspicious new packages. The risk is mostly theoretical right now, but researchers recommend acting early to prevent future problems.

Anthropic, others detail how to build reliable AI workflows
AI Deep Dives & Tutorials

Anthropic, others detail how to build reliable AI workflows

Reliable AI workflows may be built by treating each step as a clear, dependable product rather than a loose group of scripts. Experts suggest using modular pipelines, making each part deterministic and observable, and adding complexity only when needed. Regular checks, retries, and sometimes human review should be included to handle failures or uncertain results. Monitoring tools and clear targets for data quality, model output, and speed help teams notice problems quickly and decide when to stop or fix issues. Tools like Airflow, Prefect, and Kubeflow each offer different ways to manage and track these workflows, but all should keep detailed logs and version control for easier troubleshooting.

Veracode: 45% of AI-generated code snippets contain security flaws
AI News & Trends

Veracode: 45% of AI-generated code snippets contain security flaws

Recent research suggests that 45% of AI-generated code snippets may contain security flaws. While AI tools can help teams create features quickly, proving that the code is safe and reliable often takes much longer. Some studies show that AI assistance leads to more completed tasks without lowering code quality, but others report more bugs and duplicated code. Experts recommend treating all AI-generated code as untrusted and using careful review and testing to find problems. Many teams are now adding security checks early, having experts review risky changes, and being careful about where and how they use AI in critical systems.

New AI Metric Tracks Tokens Per Shipped Product
AI News & Trends

New AI Metric Tracks Tokens Per Shipped Product

A new project proposes collecting data on how many AI tokens are used per shipped product across different companies, which may help measure AI efficiency better than just counting tokens. Experts suggest tracking tokens per completed task or feature, since raw token numbers may not show real business impact. The project would use privacy-safe, aggregated data and focus on metrics like tokens used, tasks finished, and defect rates. Current studies suggest that many AI projects do not show clear financial gains, so linking token use to shipped results might help spot effective practices. If done carefully, this approach could let teams and leaders compare AI efficiency more accurately.

Anthropic unveils Claude Opus 4.8: Faster, cheaper, safer AI
AI News & Trends

Anthropic unveils Claude Opus 4.8: Faster, cheaper, safer AI

Anthropic has released Claude Opus 4.8, which may be faster, cheaper, and safer than its earlier version. The company says Opus 4.8 costs about one-third as much and gives responses around 2.5 times quicker than Opus 4.7 in fast mode. Early tests suggest it performs well at fixing complex code and avoids misaligned behavior more often, but some engineers might still prefer other tools like Codex for certain quick tasks. Anthropic's own checks indicate the new model flags uncertainty more and avoids unsupported claims, though these results may need confirmation from outside labs. Pricing stays similar to before, and actual savings might depend on how teams use it.

AWS Trainium Cuts AI Costs Up to 50% for Anthropic, Uber
AI News & Trends

AWS Trainium Cuts AI Costs Up to 50% for Anthropic, Uber

Amazon, Google, and Meta are now competing in AI by focusing on things like computer chips, data systems, and how companies make money from AI. AWS's custom Trainium chips reportedly help companies like Anthropic and Uber cut AI costs by up to 50 percent, though the use of these chips is still selective. Microsoft is focusing on controlling data access and security, which might make it harder for companies to switch away from their products. Reports suggest most AI startup revenue is now going to just two companies, OpenAI and Anthropic, which together may control about 89 percent of the market. These trends suggest that control over chips, data, and APIs may decide which companies lead in AI, but it is not yet clear who will win.

AI Code is 86% Flawed in Benchmarks, CSET Study Finds
AI News & Trends

AI Code is 86% Flawed in Benchmarks, CSET Study Finds

A recent study suggests that AI-generated code has many flaws, with failure rates as high as 86% for some security issues. Researchers found that nearly half of AI code samples show problems like SQL injection and improper authentication. The main challenge may be checking and fixing AI code, not just creating it, so teams are adding more human reviews and strict checks. Productivity gains are mixed; AI might speed up some tasks, but could also slow down experienced programmers and increase maintenance work. Experts recommend careful oversight and collecting data to ensure code quality and safety.

AI hallucinations now enable 'slopsquatting' attacks, say security experts
AI News & Trends

AI hallucinations now enable 'slopsquatting' attacks, say security experts

Security experts say AI tools sometimes invent fake package names, which attackers might quickly register to trick automated coding systems. These so-called "slopsquatting" attacks may allow bad actors to silently add malicious software when teams use AI to pick code dependencies. Tests in 2025 and 2026 show that many fake names created by AI were actually downloaded or used widely, even though they did not include harmful code. Experts suggest multiple security steps may reduce, but not fully stop, these risks. It remains unclear how often these attacks lead to real harm, since there is little direct evidence so far.

AI-generated code triples remediation time, costs $4.88M per breach
AI News & Trends

AI-generated code triples remediation time, costs $4.88M per breach

AI-generated code may create security gaps that are hard to spot and fix. When problems happen, fixing the damage reportedly takes about three times longer than with human-written code and costs around $4.88 million per breach on average. Studies suggest that almost half of AI-generated code samples may have security flaws, and big companies have seen a big rise in security issues after using AI tools. Experts recommend strict access controls and regular checks to lower risks, but teams often spend a lot of time just figuring out what the AI was supposed to do before fixing any problems.

AI Coding Agents Shift Engineers to Oversight in 2026
AI News & Trends

AI Coding Agents Shift Engineers to Oversight in 2026

By mid-2026, coding agents may be handling most of the routine coding work, such as writing code, testing, and making pull requests, while human engineers focus on supervision and review. Reports suggest agents like OpenAI Codex, Claude Code, Cursor Composer, and GitHub Copilot Agent Mode now manage full workflows, including debugging and documentation. Teams choose between using one agent for simple tasks or several agents in parallel for bigger, modular tasks, depending on the job and oversight needs. No single tool appears to lead in all benchmarks, and the best option seems to depend on the kind of task. Forecasts suggest engineers might spend more time on oversight and coordination, though some experts warn that progress may be slower than hoped.