Content Fans

Category

AI News & Trends

1402 articles • Page 62 of 94

Sierra raises $950M, powers 40% of Fortune 50 voice agents

Sierra raises $950M, powers 40% of Fortune 50 voice agents

Sierra may be becoming a key player in the voice AI market after raising $950 million and now serving 40 percent of Fortune 50 companies. The company's revenue recently grew to over $165 million, but the wider market is still considered early stage. Research suggests many big companies are just starting to use voice agents, with only about two-thirds of Fortune 500 firms having these systems. While Sierra's growth suggests full adoption may be beginning, experts say challenges like regulation and security remain. The new funding might help Sierra improve its platform as more businesses consider voice agents for customer service.

AI Tool Poisoning: Hackers Exfiltrate Data From Assistants in New Supply Chain Attack

AI Tool Poisoning: Hackers Exfiltrate Data From Assistants in New Supply Chain Attack

Hackers may be able to trick AI assistants into sending private data by hiding instructions in app descriptions, a method called AI tool poisoning. This attack appears to work across different assistants like ChatGPT and Claude, because they trust and follow hidden commands in tool metadata. Security experts found that these attacks succeed over 60 percent of the time, and may have already caused many costly data breaches. Defending against this is hard because the attack hides in normal workflows, and retraining models does not always fix it. Experts suggest using signed tools, sandboxing, and human approval, but a full solution may not be available yet.

Google Zero: Publishers lose 33% traffic as AI Overviews expand

Google Zero: Publishers lose 33% traffic as AI Overviews expand

Publishers may be losing a lot of website traffic from Google because of new AI Overviews in search results. Reports suggest website visits dropped by about 33 percent worldwide, possibly due to people getting answers directly on Google without clicking links. Experts say this change appears to hurt sites with general or repeated content the most, while those with unique, original reporting or loyal audiences are less affected. Publishers might be shifting focus to things they control, like newsletters and direct communities, and looking for new ways to earn money beyond pageviews. There is still uncertainty about the best strategies, but building original and distinctive content seems to help.

Anthropic launches Claude Platform on AWS with full feature parity

Anthropic launches Claude Platform on AWS with full feature parity

Anthropic has made its Claude Platform on AWS available, allowing developers to use the Claude Messages API with their AWS accounts. The platform may offer all the same features and pricing as Anthropic 's direct API, but data appears to be processed outside AWS, unlike Amazon Bedrock. Both services use the same Claude models, but new features might appear on the Claude Platform faster than Bedrock. Companies concerned about strict data rules may still need Bedrock. Analysts suggest AWS might be trying to attract both fast-moving teams and highly regulated users with these options.

GameDiscoverCo unveils 5 factors for predicting Steam game success

GameDiscoverCo unveils 5 factors for predicting Steam game success

GameDiscoverCo suggests that predicting Steam game success may work better with a model that uses several signals instead of just wishlists. They point to five factors: demo engagement, Discord or forum activity, influencer reach, follower-to-wishlist ratio, and wishlist velocity. Each factor is adjusted based on the game's genre and audience. The final prediction appears as a probability range, such as a 30-45 percent chance of reaching 100,000 sales in the first month. However, results may still be uncertain due to outliers and unique events.

UBS Names Microsoft, OpenAI, Nvidia "Key Enterprise AI Winners"

UBS Names Microsoft, OpenAI, Nvidia "Key Enterprise AI Winners"

UBS says Microsoft, Nvidia, and OpenAI may be the main winners in enterprise AI, based on a survey of IT leaders and new analyst ratings. The report suggests that more companies are using AI, with Nvidia chips and Microsoft Azure leading in their areas, while OpenAI's models seem to be the top choice. UBS analysts believe these three firms work well together for businesses, making it easier for companies to use AI from hardware to apps. The report also notes other companies might benefit, but leadership appears to stay with Microsoft, Nvidia, and OpenAI for now, although this could change if new technology costs drop quickly.

Microsoft's Azure GPU Policy Shifts, Pressures OpenAI's Hardware Plans

Microsoft's Azure GPU Policy Shifts, Pressures OpenAI's Hardware Plans

Microsoft appears to be making it harder for smaller customers, including OpenAI, to get high-end GPUs on its Azure cloud due to supply shortages. The company now gives priority to its top spenders and internal projects, meaning many smaller firms face long waits or quota denials. OpenAI still depends on Azure for much of its AI work, but may need to speed up building its own chips and data centers or look at other cloud options. These challenges may continue through at least late 2026, and could slow OpenAI's plans to become less reliant on Microsoft and Nvidia.

American Express Unveils AI Payments with ACE Developer Kit

American Express Unveils AI Payments with ACE Developer Kit

American Express has introduced the ACE Developer Kit, which aims to let AI agents make payments using intent contracts and single-use tokens. This system may give American Express more control and could help prevent mistakes or fraud in AI-led purchases. Three parts of the kit - Account Enablement, Intent Intelligence, and Payment Credentials - are already available, while others are still being developed. Regulators and competitors appear to be watching closely, as this new way of handling AI payments might shape future industry standards. The company suggests that more features and updates are expected in 2026, but the program is still early in its rollout.

Anthropic Launches Claude Managed Agents in Public Beta

Anthropic Launches Claude Managed Agents in Public Beta

Anthropic launched Claude Managed Agents in public beta in April 2026. This service may let teams run agents easily by handling session memory and orchestration on Anthropic's servers, so users can focus on designing prompts. Pricing stays pay-as-you-go, with standard token rates and a session hour fee, and usage costs might range from $10 to $150 per month depending on the workload. Early reports suggest companies like Rakuten quickly set up agents for different departments, and some legal and financial teams appear to have sped up their work using these agents. Tests suggest Claude Managed Agents may cost more than some competitors but offers detailed billing and large context windows.

SAP Unveils Autonomous Enterprise Suite, AI Agents at Sapphire 2026

SAP Unveils Autonomous Enterprise Suite, AI Agents at Sapphire 2026

SAP has announced the Autonomous Enterprise suite, which may help businesses automate tasks using AI agents and a unified platform. The suite combines several SAP technologies and includes over 50 specialized assistants to manage business processes. Early versions are expected to launch in finance, planning, HR, and customer experience areas throughout 2026. Pricing details are not fully shared, and access is bundled through SAP's cloud programs. SAP suggests that the new system aims to improve efficiency and maintain governance, but some results, like faster financial close cycles, are described as goals rather than guarantees.

ParcelBio Unveils $13M Seed Round for Next-Gen mRNA Platform

ParcelBio Unveils $13M Seed Round for Next-Gen mRNA Platform

ParcelBio has raised $13 million to develop its new mRNA platform, which may help make protein treatments last longer in the body. Other companies also recently raised funds for similar next-generation mRNA and RNA therapies that might work better or reach more tissues. The science behind these therapies suggests they can be made stronger or last longer using special RNA designs. The US FDA recently introduced new ways that could help these products get approved faster, but experts say safety and manufacturing still need careful review. Analysts suggest that the first long-term mRNA treatments may be approved for use as early as 2028, though this depends on continued positive safety results.

OpenAI open sources Agentic Commerce Protocol for AI shopping

OpenAI open sources Agentic Commerce Protocol for AI shopping

The Agentic Commerce Protocol (ACP) is an open standard from OpenAI that lets AI handle online shopping tasks like searching products, building carts, and processing payments, while merchants keep control. ACP mainly covers the customer-facing parts of shopping, and leaves things like inventory management to other systems. As of early 2026, adoption appears limited, with only a small number of merchants using ACP checkouts, possibly due to lower conversion rates compared to traditional methods. New standards like Google's Universal Commerce Protocol may increase competition, and it seems likely that multiple protocols will be used instead of just one. Merchants are experimenting with ACP to learn how AI can recommend products, and ongoing changes may affect how AI shopping works in the future.

Far Far West Uses Playtests to Trigger Steam Discovery Algorithm

Far Far West Uses Playtests to Trigger Steam Discovery Algorithm

Far Far West may have triggered Steam's Discovery Queue in early 2025 by running two tightly scheduled public playtests and using creator streams. These events created short bursts of engagement, like high player numbers and many wishlists, which seem to be important for Steam's recommendation system. Developers report that these concentrated playtests can help games appear in "Because you played" and "More Like This" sections, especially when paired with positive ratings and creator coverage. Recent changes to Steam's algorithm appear to favor sales and wishlists over tag similarity, making these activity spikes more important for visibility. This strategy might keep working as long as Steam's algorithm continues to value short-term engagement signals.

CF

AI FinOps Tools Emerge to Track Escalating LLM Costs

AI FinOps tools are being developed to help companies track and manage the rising costs of using large language models (LLMs). Reports suggest that many businesses are surprised by LLM expenses and want better ways to see and control these costs, similar to cloud services. Surveys show most AI startups now spend more on running models than training them, and almost all companies track their AI spending. New tools and dashboards may help companies forecast costs, spot savings, and link spending to business value, but there are still challenges like missed forecasts and gaps in AI governance. Early evidence suggests these tools might help reduce cost surprises and start important budget discussions sooner.

OpenAI, Microsoft, Google Ship Enterprise AI Agents Ahead of EU Act

OpenAI, Microsoft, Google Ship Enterprise AI Agents Ahead of EU Act

OpenAI, Microsoft, and Google have each released new AI agents for businesses, which may signal a shift in how companies use AI in daily work. These agents are designed to run tasks automatically, share progress, and follow rules for safety and privacy. Google's and Microsoft's tools appear to help teams automate tasks and handle routine jobs, but setup and costs might vary. The European Union has set August 2, 2026, as a key deadline for new AI rules, suggesting that companies must make sure their AI systems are well-documented and follow strict controls. This may mean that both speed in creating AI and following new laws will become very important for companies in the coming year.