IBM says company is investing in products it can use: We drink our own… – The Times of India
IBM is putting its venture dollars where its enterprise roadmap is headed. Emily Fontaine, who leads IBM Ventures, says the $500 million fund is concentrating on startups whose products IBM can immediately deploy across its own operations—making the tech giant “client zero” for many portfolio companies. The result is a tight loop between investing, co-building, and real-world validation, particularly in artificial intelligence and quantum computing.
IBM’s inside-out thesis: build, buy, then be the first user
Fontaine describes the strategy succinctly: “We drink our own champagne,” meaning IBM aims to implement the capabilities it backs, fast. One internal showcase is AskHR, an AI-driven human resources assistant used by IBM’s workforce. Instead of routing routine queries through managers and HR staff—everything from relocation help to mortgage rate guidance—employees tap AskHR to get information instantly. That shift illustrates how IBM evaluates startups: can this technology ship, scale, and remove friction inside a complex enterprise?
Focus areas: B2B AI and quantum that map to IBM’s client base
IBM Ventures has made 23 investments to date, prioritizing B2B startups that align with the company’s core markets and long-standing customers. Recent bets span:
- Hugging Face: tooling and infrastructure for machine learning developers.
- Not Diamond: model-selection optimization to match AI models to specific tasks.
- Unstructured: data preparation pipelines for large AI models.
- Reality Defender: deepfake detection and media authenticity safeguards.
- QEDMA: quantum software focused on error mitigation for noisy quantum hardware.
That portfolio mix underscores an emphasis on practical AI adoption—making it easier to train, select, and govern models at scale—while laying groundwork for the quantum era.
The “capital-plus” model: investment, distribution, and collaboration
Fontaine frames IBM Ventures as more than a checkbook. The “plus” is access to IBM’s enterprise channels, co-selling muscle, and technical integration pathways—support that most startups struggle to build quickly. She says more than 90% of portfolio companies collaborate directly with IBM teams and clients.
Every investment is assessed across three criteria:
- Products or capabilities IBM can use or embed.
- Ecosystem fit as a partner to IBM platforms and services.
- Potential to disrupt industries, often in close coordination with IBM Research.
For startups, that means a structured route to enterprise validation; for IBM, it means earlier access to differentiated tech that can be operationalized safely inside regulated environments.
AI payoffs: efficiency, governance, and security
Internally, IBM expects to save around $4.5 billion this year as AI tools streamline workflows. That figure reflects both automation and augmentation—using AI to accelerate processes rather than merely reduce headcount. Crucially, Fontaine stresses “responsible AI” as an investment filter, reflecting client demand for robust governance, model provenance, and risk controls.
From a cybersecurity standpoint, the inclusion of Reality Defender signals that AI trust and safety is now a board-level concern. Deepfake detection, content authenticity, and resilient model operations are increasingly prerequisites for enterprise AI rollouts, not optional add-ons. By prioritizing vendors that harden the AI supply chain—data prep, model selection, and verification—IBM is betting that security-by-design will differentiate winners in B2B AI.
Quantum strategy: software-first, security-forward
While IBM has built and tested its own quantum chips, Fontaine says the venture focus has been squarely on software and algorithms—the layers most likely to deliver near-term value as hardware matures. QEDMA, an Israeli startup, is emblematic: it tackles error mitigation for today’s noisy quantum systems, a practical bottleneck for running useful workloads.
Financial services is a major driver. Banks are pressing for “quantum-safe” strategies amid concerns that future quantum machines could break widely used public-key cryptography. That urgency is spurring planning around post-quantum cryptography (PQC), crypto-agility, and phased migrations—projects that demand years of testing and inventorying before full switchover.
For CISOs and security architects, the message is clear: start the quantum risk program now. Establish cryptographic inventories, evaluate NIST-selected PQC algorithms, and design migration paths before regulatory or competitive pressures force rushed transitions.
Exits and returns: progress without specifics
IBM hasn’t disclosed overall returns for the fund, though Fontaine says she is satisfied with performance so far. Four startups have already exited, including two notable acquisitions: Wiz’s purchase of Gem Security for a reported $350 million and Cisco’s acquisition of Lightspin for an estimated $200–$250 million. The activity suggests healthy demand for modern cloud security and infrastructure tooling—the very categories IBM’s clients continue to prioritize.
Why this matters for enterprises
IBM’s “client zero” approach compresses the gap between innovation and reliable enterprise deployment. For CIOs and CISOs, that can translate to:
- Faster proofs of value: technologies are vetted at IBM scale before hitting broader customer environments.
- Stronger security posture: emphasis on responsible AI, deepfake detection, and quantum-safe planning aligns with regulatory and risk mandates.
- Reduced integration risk: portfolio startups are primed for IBM’s platforms and partner ecosystem.
In an era of AI hype and quantum uncertainty, IBM Ventures is signaling a pragmatic path: invest in tools the enterprise will actually use, prove them internally, then scale through established channels. If the reported cost savings hold and quantum-readiness efforts continue to accelerate, the “capital-plus” playbook could become a template for how large tech companies de-risk next-gen capabilities for the Fortune 500.
