OpenAI, Anthropic, AWS, Microsoft 100+ companies warn of AI-enabled cyberattacks, urge action

There’s a “limited window” to strengthen cyber defenses, the companies said.

Hospitals, water treatment plants, and other critical infrastructure are particularly vulnerable to AI-enabled attacks.

Organizations should raise internal security standards and fix their “highest-risk weaknesses,” the companies said.

The signatories made no specific commitments, deadlines or investment pledges as part of the initiative.

Most fraud programs are built around decision points, but fraud moves across accounts, channels, and organizations long before it reaches your systems.

In 2025, those blind spots contributed to $38 billion in identity fraud and scam losses. Staying ahead means catching risk earlier with more connected data.

In this report, Javelin breaks down the four fraud trends you should keep an eye on—and how to spot risk before it hits your bottom line.

Last week: NVIDIA stepped back from some deals under the revenue-sharing program launched in Jul.

The program could draw antitrust scrutiny, some NVIDIA employees raised concerns with current and potential customers.

NVIDIA could revamp the program or fold it into another initiative.

The program is “still in place and continues to evolve due to high demand,” an NVIDIA spokesperson said.

Your team probably already uses a handful of different AI tools. This piece rounds up what's new in Slackbot and how it pulls all of them into one place.

See the latest ways Slackbot helps you get answers, draft work, and move projects forward, without bouncing between apps to make your AI tools talk to each other.

Whether you're just getting started with AI or already have a few tools in rotation, get a clear look at what Slackbot can do for your team right now.

“The empty invocation of national security is not a blank check to punish and retaliate against government critics,” U.S. District Judge Rita Lin wrote.

Anthropic welcomed the ruling, saying it remained “focused on working productively with the government to harness AI… so all Americans benefit from this technology.”

Separately, Anthropic released Model Hardware Standard to help AI agents operate microscopes, quantum hardware, and robot arms.

Anthropic also reportedly plans to allow secondary stock sales in its IPO while weighing longer-than-standard lockups.

Audio transcription and speech data labeling

Image and video annotation for vision model pipelines

Multilingual coverage across Portuguese, Spanish, and English

Advent and Stripe could return with an offer later if circumstances change.

PayPal stock -14% in pre-market trading on Fri. on the news.