IBM shares plunge 25% as customers shift spending to AI
IBM shares fell 25% as CEO Arvind Krishna stated customers shifted spending from traditional IBM offerings to AI infrastructure.
Financial Times Technology·14 Jul
Search signals, briefings, company results, benchmarks and glossary terms.
Search signals, briefings, company results, benchmarks and glossary terms.
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
IBM shares fell 25% as CEO Arvind Krishna stated customers shifted spending from traditional IBM offerings to AI infrastructure.
Anthropic's latest marketing campaign, featuring an ad critics find unsettling, leverages AI's negative perceptions to highlight its safety focus.
SpaceXAI's Grok Build AI coding tool was caught uploading users' entire codebases, including excluded files, to Google Cloud storage.
Thrad.ai deployed a multi-agent system on AWS Bedrock for prospect discovery and email generation, benchmarking Swarm vs. Graph orchestration.
Major publishers Hachette, Cengage, and Elsevier are suing Google, alleging its AI models were trained on copyrighted works without permission.
DeepMind CEO Demis Hassabis proposes an independent AI standards body, modeled on FINRA, to evaluate frontier models and set release best practices.
Former Meta employees are suing, alleging the company used AI tools to unfairly target workers on leave for mass layoffs based on performance data.
AWS details batch regression testing and CI/CD integration for agentic QA automation using Amazon Nova Act, extending prior capabilities.
AWS introduced Amazon Nova Act, a generative AI solution for automated, large-scale user experience testing, capable of generating scenarios and analyzing user flows.
Flo Health built a production-grade medical content review and generation system using Amazon Bedrock, scaling from a POC.
ScienceSoft leveraged Amazon Bedrock Guardrails to build a HIPAA-compliant AI voice scheduler using Amazon Nova 2 Sonic.
Meta's Adam Mosseri suggested companies may soon cap AI token spending per engineer to manage costs, similar to other operating expenses.
Spotify is piloting a new AI chatbot feature for Premium subscribers, enabling conversational search and discovery of music and audio content.
New York State has imposed a temporary halt on new large data center construction due to concerns over electricity costs, water supply, and local control.
FBI warns again about continued Russian state-sponsored attacks targeting poorly configured routers and Cisco devices in critical infrastructure.
Menlo Ventures discusses the emerging architectural shift in cybersecurity, advocating for AI-native defense to counter automated machine-speed attacks.
Reflection AI, an open-source AI startup founded in 2024, secured a $1 billion compute deal with Nebius.
Grok Build was alleged to have uploaded users' entire codebases to SpaceXAI without explicit user consent, despite Musk's privacy claims.
The FCA proposed reforms for asset managers, aiming to cut £128m annually through simpler Fund Reporting for Asset Management Entities (FRAME) requirements.
Hugging Face CEO Clem Delangue claims enterprises increasingly prefer open models for cost, accessibility, and ownership over frontier models.
Insight Partners observes a shift in enterprise AI adoption where clients prioritize solving specific problems over platform-first strategies, leveraging 'forward deployed engineers'.
Spotify is rolling out an AI-powered conversational assistant to its Premium subscribers for music, podcast, and audiobook discovery.
Apple has filed a high-profile lawsuit against OpenAI, challenging the company's hardware bets and increasing its existing legal entanglements.
Superhuman's new auto-draft feature for email replies is reported to generate highly accurate suggestions requiring minimal editing, enhancing productivity.
Buildkite is gaining traction as a CI/CD platform for AI development; Meta invests in compute infrastructure.
Cloudflare deployed a Negative Trust Anchor after a DNSSEC key rollover failed for the .al TLD, and 1.1.1.1 now returns EDE 33 to signal DNSSEC bypass.
1Password launched AI Spend and Consumption Management, a new capability in its SaaS Manager platform for tracking AI service costs.
CISA has added four vulnerabilities, including flaws in SonicWall SMA1000 and Microsoft products, to its KEV Catalog due to active exploitation.
CISA released an advisory on a critical vulnerability in Rockwell Automation 1715-AENTR EtherNet/IP Adapter, allowing file manipulation and I/O state changes.
CISA warns of active exploitation of vulnerabilities (CVE-2026-32201, CVE-2026-45659, CVE-2026-56164) in on-premises SharePoint Servers.
ABB is addressing a Linux kernel vulnerability (CVE-2026-31431) in its Edgenius product allowing local privilege escalation to root.
DeepMind chief Demis Hassabis called for a US-led body to test 'frontier' AI models, citing a 'precious window' to prepare for rapid advancements.
Google DeepMind CEO Demis Hassabis advocates for a global AI watchdog led by the US to regulate frontier models and prevent potential dangers.
Chinese AI startup DeepSeek is seeking new funding rounds shortly after closing its initial round, indicating rapid infrastructure expansion needs.
OpenAI published guidance on managing AI investments in the 'agentic era', focusing on measuring useful work per dollar and scaling high-value workflows.
The 1960s chatbot ELIZA foreshadowed user willingness to confide in AI, setting a precedent for current LLM privacy concerns.
BIS paper warns that the AI build-out, driven by debt and circular equity, risks over-investment and potential bust due to revenue disappointment.
New York enacted the first statewide moratorium on new hyperscale data centers, blocking environmental permits for up to one year.
HUD is withholding documents about the use of AI by DOGE in housing policy, citing a non-existent privilege.
SoftBank CEO Masayoshi Son criticizes business leaders wary of AI adoption, urging them to fully embrace new technologies.
US export controls on advanced AI chips are fueling a black market in China, impacting global semiconductor supply chains.
Financial Times reports on AI's potential to facilitate new forms of deception and fraud, drawing parallels to the 'Mechanical Turk' illusion.
AI lawsuits are exposing gaps in conventional insurance coverage, particularly with the shift from chatbot errors to autonomous agent liabilities.
Research demonstrates an imputation-free transformer model improving Alzheimer's prediction and uncertainty quantification from incomplete clinical data.
Geometric Causal Models (GCMs) are a new framework for causal inference from dependent data, such as spatial or network data, using symmetries.
Research introduces Legendre Jumper Martingales for online detection of distribution shifts, framing it as an exchangeability violation.
Research proposes Propagational Proxy Voting (PPV) for multi-sample LLM inference, aggregating by delegation using entropy-based confidence and reasoning-embedding similarity.
Researchers developed a branched signature kernel solver for ODEs driven by single, rough forcing signals, applicable in finance and engineering.
A research survey identifies statistical challenges hindering reinforcement learning (RL) deployment in real-world settings despite its successes.
Research introduces SPQR, a benchmark for evaluating the stability of safety alignment in text-to-image diffusion models under benign fine-tuning, observing frequent breakdowns.
© 2026 OneBench: AI Insights. All rights reserved.
Evidence before opinion