Yesterday, Tom Tunguz reported $10B in forward-deployed engineering commitments, signaling a new front in enterprise AI: companies embed engineers to build moats even as AI-powered attacks rise.
Yesterday, investor Tom Tunguz spotlighted a staggering statistic: AI companies have collectively committed $10 billion to forward-deployed engineering this year. This isn’t a bug fix—it’s a strategic moat. By embedding engineers directly inside customer organizations, AI firms are locking in deployment, managing inflection points, and capturing feedback loops that pure SaaS can’t touch. The revelation, published on 07 July 2026, crystallizes a whirlwind month in enterprise tech.
At SaaStr AI 2026, also on 07 July, the data confirmed the shift: total software spend jumped 15% year-over-year, hitting $1.4 trillion, but the gains are uneven. AI-powered platforms are soaring while legacy SaaS valuations crater. The B2B software market is bifurcating in real time—AI adopters thrive, laggards stall.
Two weeks ago, on 24 June 2026, Glean CISO Sunil Agrawal detailed a darker trend: AI-generated phishing and reconnaissance attacks that lack traditional tells like typos or suspicious headers. These AI-native threats are bypassing signature-based defenses, forcing a top-to-bottom revision of cybersecurity. CrowdStrike’s 2026 Threat Report, released in June, corroborates the uptick.
This month, the talent war for forward-deployed engineers intensified. LinkedIn’s Workforce Report showed a 42x spike in job postings for such roles in 2026, as companies scramble to place AI experts on the ground. The convergence is unmistakable: CIOs must fight on two fronts simultaneously—embed AI deeply to compete, yet secure it aggressively against AI-powered intrusions.
This isn’t the first time deployment models have altered enterprise spending. In the early 2000s, the rise of SaaS was itself a deployment revolution, shifting software from on-premise installations to cloud subscriptions. Today’s forward-deployed engineering echoes the professional services swell of the ERP era, when consultants camped on site for years—but now AI agents augment the work at blazing speed.
Cybersecurity has also seen paradigm shifts before: from antivirus in the 1990s to behavioral analytics in the 2010s. Yet the AI-native attacks reported this month are qualitatively different—they adapt in real time, learn from defenses, and exploit human psychology without a hint of tradition. As of 08 July 2026, the battle lines are drawn: enterprises that don’t embed AI lose the growth race; those that embed without securing it risk catastrophic breaches.