Mythos Drove Up Cyber Risk 10-Fold, Says JPMorgan's Dimon
THE SO WHAT
If a single frontier model can move a Tier 1 bank CEO to claim a 10x cyber risk jump, boards are going to start treating advanced AI as both tool and threat surface. CISOs should assume adversaries are already weaponizing Mythos-class models and accelerate red-teaming, model-aware detection, and incident playbooks accordingly.
READ THE SOURCE
MORE FROM THE WIRE
Applied AILibreOffice says ‘no AI’ is now a software feature
A credible "no AI" stance turns privacy and determinism into product features, not just legal boilerplate. If you sell into governments, regulated industries, or privacy-first markets, expect RFPs to start asking explicitly whether AI is on by default and where data goes.
Applied AISAP to acquire AI startup TechWolf for undisclosed sum
SAP buying TechWolf is a bet that skills intelligence becomes a core ERP primitive, not a bolt-on HR feature. If you’re selling into large enterprises with your own talent graph or workforce AI, assume SAP will try to bundle this into every HXM deal and compress your room to maneuver.
Applied AIAnthropic earmarks $100 million to train over 10,000 AI engineers in push to develop enterprise AI
Putting $100M into training 10,000 external AI engineers is a distribution play—Anthropic is trying to make Claude the default in enterprise build-outs by seeding talent. If you’re an enterprise leader, expect vendor-backed "certified" AI talent pipelines to show up in your org and decide now how much to lean into one ecosystem.
Applied AIAI to See Massive Adoption 'Unseen' in Previous Tech Cycles, Says Neostellar Principal
If US–China AI performance gaps are narrowing as labs like DeepSeek catch up, the assumption of a durable Western technical lead becomes less safe. For operators, that means planning for a multipolar AI supply chain—on models, talent, and infra—rather than betting on a single geography.