InfoWorld

Is your generative AI application giving the responses you expect? Are there less expensive large language models —or even free ones you can run locally —that might work well enough for some of your tasks? Answering questions like these isn’t always easy. Model capabilities seem to change every month. And, unlike conventional computer code, LLMs don’t always give the same answer twice. Running an…

We just can’t seem to help ourselves. Our current infatuation with multi-agent systems risks mistaking a useful pattern for an inevitable future, just as we once did with microservices . Remember those? For some good (and bad) reasons, we took workable applications, broke them into a confusing cloud of services, and then built service meshes, tracing stacks, and platform teams just to manage the …

Databricks has outlined a reference architecture for a multi-agent AI system, named AiChemy, that combines internal enterprise data on its platform with external scientific databases via the Model Context Protocol (MCP) to accelerate drug discovery tasks such as target identification and candidate evaluation. These early-stage steps are critical in drug development because they help pharma compan…

Car buyers kick tires. Horse traders inspect the teeth. What should shoppers for large language models (LLMs) do? Here are 27 prescient questions that developers are asking before they adopt a particular model. Model capabilities are diverse, and not every application requires the same support. These questions will help you identify the best models for your job. What is the size of the model? The…

Anthropic has blocked paid Claude subscribers from using the widely used open-source AI agent OpenClaw under their existing subscription plans, a move that took effect April 4 and has drawn pushback from subscribers who question both the cost implications and the company’s stated rationale. In an email to subscribers reviewed by InfoWorld, Anthropic said access to third-party tools through subscr…

Researchers who identify and report bugs in open-source software will no longer be rewarded by the Internet Bug Bounty team. HackerOne, which administers the program, has said that it is “pausing submissions” while it contemplates ways in which open source security can be handled more effectively. The Internet Bug Bounty program, funded by a number of leading software companies, has been run sinc…

The leak of Claude Code’s source is already having consequences for the tool’s security . Researchers have spotted a vulnerability documented in the code. The vulnerability, revealed by AI security company Adversa , is that if Claude Code is presented with a command composed of more than 50 subcommands, then for subcommands after the 50th it will override compute-intensive security analysis that …

The European Union’s Computer Emergency Response Team, CERT-EU, has traced last week’s theft of data from the Europa.eu platform to the recent supply chain attack on Aqua Security’s Trivy open-source vulnerability scanner. The attack on the AWS cloud infrastructure hosting the Europa.eu web hub on March 24 resulted in the theft of 350 GB of data (91.7 GB compressed), including personal names, ema…

Google has added two new service tiers to the Gemini API that enable enterprise developers to control the cost and reliability of AI inference depending on how time-sensitive a given workload is. While the cost of training large language models for artificial intelligence has been a concern in the past, the focus of attention is increasingly moving to inferencing , or the cost of using those mode…

Let’s begin with the core question: Is OpenClaw a cloud entity or not? The best answer is a complicated “not exactly, but functionally, yes.” OpenClaw AI Agent Platform is better viewed as an orchestration layer, runtime, or plumbing rather than a complete cloud platform. It provides the tools to build and manage agents but lacks the intelligence, data estate, control plane, or business context t…

If JavaScript were a character in a role-playing game, its class would be a Rogue. When it was a youngster, it was a street kid that lived on the margins of society. Over time, it has become an established figure in the enterprise hierarchy. But it never forgot where it came from, and you never know what sleight of hand it will perform next. For example, fine-grained Signals are mounting a rebell…

Anthropic likes to talk about safety. It even risked the ire of the US Department of Defense (also known as the Department of War) over it. But two unrelated leaks in the space of a week have put the company in an unfamiliar spotlight: not highlighting model performance or safety claims, but for its apparent difficulty in keeping sensitive parts of its AI tooling and strategy out of public view. …

Kilo has launched KiloClaw for Organizations, a managed version of its OpenClaw platform aimed at enterprises seeking more control over how employees deploy AI agents for tasks such as repository monitoring, email drafting, and calendar management. Co-founded by GitLab co-founder Sid Sijbrandij and Scott Breitenother, Kilo is building open-source coding and AI agent tools and is gaining attention…

Artificial intelligence and related technologies are evolving rapidly, but until recently, Java developers had few options for integrating AI capabilities directly into Spring-based applications . Spring AI changes that by leveraging familiar Spring conventions such as dependency injection and the configuration-first philosophy in a modern AI development framework. My last tutorial demonstrated h…

Computer vision teams face an uncomfortable reality. Even as annotation costs continue to rise, research consistently shows that teams annotate far more data than they actually need. Sometimes teams annotate the wrong data entirely, contributing little to model improvements. In fact, by some estimates, 95% of data annotations go to waste . The problem extends beyond cost. As I explored in my prev…