AI news summaries
Essential facts and source links for recent ai stories.
-
Motley Fool contributor favors Marvell over Arm on growth and valuation
A Motley Fool contributor favors Marvell over Arm in 2026, citing fiscal 2026 revenue growth of 42.1% versus 22.8% and a forward P/E of 58 times versus Arm’s 123.7. Marvell’s 10 largest customers, however, generated 82% of its fiscal 2026 revenue. -
Teradyne unveils three AI-focused test instruments; Q2 revenue rises 104%
Teradyne introduced three UltraFLEXplus instruments on Sept. 1 for AI and data-center chip testing. Q2 2026 revenue rose 104% year over year to $1.329 billion, while management projected Q3 revenue of $1.2 billion to $1.3 billion. -
Amazon monetizes AI through AWS as Alibaba builds its own AI stack
Amazon's AWS revenue grew 37% year over year to $42.2 billion in Q2; the company says its AI business topped a $25 billion annual run rate. Alibaba's AI cloud grew 45% year over year to about $7.1 billion in the June quarter, while net income fell 75%. -
Bilal Chughtai warns AI could kill humanity as Alphabet keeps investing
Former Google DeepMind engineer Bilal Chughtai says AI could kill humanity and time may be running out, urging coordination to address risks. There is no sign his warning has slowed Alphabet’s AI push; it is his assessment of a future risk. -
Palantir adds rival AI models to AIP, aiming to reduce reliance on one model
Palantir added xAI’s Grok 4.7 and OpenAI’s GPT-6 models to AIP on Sept. 24, alongside other options, pursuing a model-neutral approach meant to let eligible businesses change AI models while retaining control of data, permissions and workflows. -
Motley Fool writer favors Salesforce over Oracle as an AI stock
Motley Fool writer Chris Neiger favors Salesforce over Oracle as an enterprise AI stock, citing Salesforce’s partnerships and lower AI spending. Oracle’s cloud-infrastructure sales rose 121% to $7.4 billion in fiscal Q1 2027, but its costly AI expansion faces project hurdles. -
Nvidia rolls out agent-safety suite; executive says it could have stopped Hugging Face breach
Nvidia released safety tools for AI agents Monday. Executive Justin Boitano said they could have stopped the Hugging Face breach if used early in frontier labs’ model evaluations. OpenShell contains agents on CPUs; Sentry can cut one off if it tries to escape. -
Musk backs 2027 AGI forecast as markets remain cautious on OpenAI
Elon Musk endorsed an X post predicting major US AI labs will reach artificial general intelligence (AGI) in 2027; it drew 1.8 million views in about three hours. Traders put the odds of an OpenAI AGI announcement before 2027 at 18% on Polymarket and 14.8% on Kalshi. -
Raspberry AI connects fashion teams across the product pipeline
Raspberry AI expanded its platform to link fashion design, merchandising, wholesale, marketing and e-commerce. CEO Cheryl Liu says teams can preview products and produce campaign images in minutes, while design judgment and physical fitting remain essential. -
Microsoft reshapes Copilot with workspaces, app-building and AI agents
Microsoft's Sept. 25 Copilot revamp created Home, Code and Autopilot workspaces, with Office integration, natural-language app building and autonomous background agents. The company also introduced FinOps controls for enterprise AI spending. -
Itron launches utility AI upgrade as quarterly revenue falls 7%
Itron launched IEE Cloud AI, an AI-enabled upgrade to its meter-data management software, on Sept. 16. Q2 2026 annual recurring revenue rose 21% to $417 million and adjusted gross margin rose 460 basis points to 41.4%, while total revenue fell 7% to $563 million. -
Certain US iPhone 15 and 16 buyers may seek up to $95 in Siri settlement
US buyers of certain iPhone 15 or 16 models bought from Apple between June 10, 2024, and March 29, 2025, may claim up to $95 per phone under Apple's $250 million Siri settlement. Claims are due Dec. 21, but payouts may be lower. -
Oracle’s AI buildout drives cloud growth but raises financing concerns
Oracle’s cloud infrastructure revenue rose 121% amid its AI buildout, while Alpha Spread put trailing 12-month free cash flow through August 2026 at negative $28.72 billion. Credit default swaps hit a record, and S&P cut Oracle’s rating to BBB, one notch above junk. -
Okta’s AI-agent prospects draw higher targets as valuation stays in focus
At Oktane 2026, Okta detailed its AI-agent security strategy as analysts raised targets. Roth Capital said early deals generated a 50%–60% uplift in deal values; management expects AI to remain immaterial in FY2027, with meaningful revenue possibly from FY2028 onward. -
More than $30 billion in AI cloud contracts lift Oracle's backlog to $664 billion
Oracle's fiscal first-quarter revenue rose 30% year over year to $19.3 billion, while over $30 billion in AI cloud contracts lifted backlog to $664 billion. Free cash flow was negative $5.4 billion, versus analysts' forecast of a $9.56 billion outflow. -
SoundHound led BigBear.ai by $25.2 million in Q2 2026 revenue
SoundHound AI reported $61.9 million in revenue for the quarter ended June 30, 2026, $25.2 million above BigBear.ai’s $36.7 million, reversing their 2024 ranking. Both companies’ Q2 revenue grew year over year, with acquisitions contributing. -
Microsoft’s reported plan would more than triple data-center capacity by 2032
Bloomberg, citing people familiar with the plans, reported that Microsoft aims to grow global data-center capacity from about 12 GW to over 38 GW by 2032; Microsoft has not confirmed the target. AI-specific capacity is expected to account for about one-third of the total. -
Nebius will raise hourly GPU prices by up to 21% on Oct. 1
Nebius will raise on-demand rates for four Nvidia GPU models by 17% to 21% on Oct. 1, including B300 from $7.85 to $9.50 an hour. CFO Dado Alonso said in August that capacity sold out as fast as it came online; existing contracts won't be repriced. -
Musk’s X and SpaceXAI move to dismiss AI antitrust claims against Apple
Elon Musk’s X Corp and SpaceXAI have moved to dismiss antitrust claims against Apple, alleging its OpenAI partnership restricted AI competition. Claims against OpenAI will continue; the filing gives no reason and does not say whether a settlement was reached. -
DeepMind chief says Gemini 4 enters early post-training as launch nears
Google DeepMind chief Koray Kavukcuoglu said Sept. 23 that Gemini 4 has entered early post-training, a phase to refine performance and reliability, as Google nears launch. The model is expected to offer more advanced AI capabilities. -
BofA raises Okta target to $220 on AI potential, keeps Neutral rating
Bank of America lifted Okta's target to $220 from $200 after Oktane, keeping a Neutral rating. It cited AI-agent security potential but said near-term forecasts are little changed; the analyst sees the upside becoming clearer in fiscal 2028. -
Palo Alto launches AI subscription to detect and remediate cyber flaws
Palo Alto Networks launched Unit 42 Continuous Frontier AI Defense, an annual service using Anthropic’s Claude Mythos, OpenAI’s GPT-5.6-Cyber and open-weight models to find, validate and remediate vulnerabilities in web apps, APIs and cloud infrastructure. -
Alibaba plans an AI model with 5–10 trillion parameters, unveils V900 chip
Alibaba plans an AI model with 5–10 trillion parameters and unveiled the Zhenwu V900 chip, described as delivering triple its predecessor’s performance. V900 mass production is expected in early 2027; Alibaba Cloud targets more than 20 gigawatts of data-center capacity by 2032. -
Microsoft unveils Code, a plain-language app builder for Copilot
Microsoft added Code to its Copilot app on Sept. 25, letting users describe an app or dashboard in plain language; it runs on the engine behind GitHub Copilot. Early access is due at month’s end, with a preview for paying subscribers later this year. -
Bill Gates calls for AI safeguards, warning misuse could cause a billion deaths
Bill Gates warned that people with malicious intent could use AI to drive events causing a billion deaths, and urged law enforcement and politicians to help define safeguards and monitoring. He made the remarks in an NBC interview scheduled to air Sept. 27. -
Microsoft redesigns Copilot to challenge Claude with coding and AI agents
Microsoft unveiled a redesigned Copilot app on Sept. 25, bringing chat and task delegation, natural-language coding, and cloud-based agents into one place, with Word, Excel and PowerPoint built in. The app is intended to challenge Claude as Microsoft seeks stronger Copilot use. -
Google tests Gemini calls to businesses as Meta expands Muse
Google is testing a Gemini feature that would call businesses to book tables, check stock or reschedule appointments, then report back to users. Google has not announced wider availability; Meta has expanded Muse into shopping, travel and grocery services. -
Synchrony teams up with OpenAI to prepare for AI-assisted shopping
Synchrony Financial wants its credit, rewards and financing offers shown as shoppers use AI assistants to compare products. Its August OpenAI partnership will include a ChatGPT plugin to search partner discounts and promotional financing.
