ML/AI Engineer | Darktrace Global Negotiation Guide

Negotiation DNA: ML/AI Engineer | Darktrace (LSE: DARK) | IC Track | Self-Learning AI Cybersecurity | Cambridge & London & San Francisco

---

Compensation Benchmarks by Region

Component	Cambridge UK (GBP)	London (GBP)	San Francisco (USD)
Base Salary	£75,000 – £115,000	£85,000 – £130,000	$175,000 – $245,000
Annual Bonus	£7,500 – £17,250	£8,500 – £19,500	$17,500 – $36,750
RSU Grant (4-yr vest)	£35,000 – £80,000	£45,000 – £100,000	$90,000 – $200,000
Signing Bonus	£5,000 – £15,000	£8,000 – £18,000	$15,000 – $40,000
Total Year-1 Comp	£104,250 – £164,750	£121,250 – £192,500	$257,500 – $396,750

Equity: Darktrace RSUs (LSE: DARK) — 4-year vesting schedule. ML/AI Engineers are among the most aggressively compensated roles at Darktrace because the entire platform's competitive moat rests on model quality. Negotiate RSU grants at the top of range — your models ARE the product.

---

Negotiation DNA

ML/AI Engineers at Darktrace are the core of the company's value proposition. You design, train, deploy, and maintain the unsupervised machine learning models that detect cyber threats in real time — the self-learning AI that gives Darktrace its name and competitive differentiation. The company is listed on the London Stock Exchange (LSE: DARK) and its market capitalization is directly tied to the quality and innovation of its AI capabilities. Your negotiation leverage is among the strongest of any role at Darktrace: without your models, there is no product. The talent market for ML/AI engineers with cybersecurity domain expertise is extremely tight, with competition from Google DeepMind, OpenAI, Anthropic, and well-funded cybersecurity companies. Darktrace's Cambridge HQ sits in one of the world's densest AI talent markets, but that also means fierce local competition for talent.

Level Mapping: ML/AI Engineer at Darktrace maps to L4–L6 ML Engineer at Google, IC4–IC6 Research Engineer at Meta, Applied Scientist II–III at Amazon, and 61–65 ML Engineer at Microsoft.

---

Darktrace SECURE AI & Agentic Identity Defender Lever

On Feb 3, 2026, Darktrace published the State of AI report — a comprehensive analysis of how autonomous AI agents are reshaping the cybersecurity threat landscape. The report documented the rapid proliferation of agentic AI systems across enterprise environments and the novel identity-based attack vectors they create: AI agents can be spoofed, manipulated, or hijacked in ways that traditional security models cannot detect. Alongside the report, Darktrace launched Darktrace / SECURE AI, a new product line that applies self-learning AI to protect AI systems themselves.

As an ML/AI Engineer candidate, you are the ultimate Agentic Identity defender. You will build the models that distinguish between legitimate and adversarial AI agent behavior — the self-learning systems that protect autonomous agents from identity-based attacks. Darktrace (LSE: DARK) is investing heavily in ML/AI talent to make SECURE AI a reality, and the State of AI report has set the strategic direction.

How to use this in negotiation: Tell the hiring team that the SECURE AI product line represents a model-architecture challenge, not just a feature. Detecting agentic identity threats requires novel unsupervised learning approaches — new anomaly detection architectures, new behavioral baselines, new adversarial robustness guarantees. Frame yourself as the ML researcher/engineer who will solve these foundational problems.

---

Global Negotiation Levers

Lever 1 — Competing Offers & AI Talent War:

"I'm evaluating offers from Google DeepMind at £120,000 base and an AI cybersecurity startup at $230,000 base. Darktrace's applied AI challenge — particularly SECURE AI — is the most compelling problem I've seen. But the ML/AI talent market is extremely competitive, and I need the total compensation to reflect that."

Lever 2 — RSU Upside & Model-Revenue Link:

"My models ARE Darktrace's product. I'd like the RSU grant increased to £85,000 / $180,000 over four years. Every improvement in detection accuracy drives customer retention and expansion revenue for LSE: DARK. My equity should reflect this direct value creation."

Lever 3 — Signing Bonus & Research Transition:

"I'm leaving behind approximately £15,000 / $35,000 in unvested equity and an active research program. A signing bonus of £15,000 / $35,000 would offset this transition cost."

Lever 4 — Compute Budget & Publication Rights:

"I need access to significant compute resources for model training and experimentation. I'd like a commitment to a dedicated GPU/TPU budget and the ability to publish non-proprietary research at top-tier venues (NeurIPS, ICML, IEEE S&P). This is essential for my career trajectory and also positions Darktrace as an AI research leader."

---

Negotiate Up Strategy: For Cambridge, target £105,000 base with £70,000 in RSUs (4-yr) and a £12,000 signing bonus. For London, push for £120,000 base with £85,000 in RSUs and £15,000 signing bonus. For San Francisco, anchor at $230,000 base with $180,000 in RSUs and $35,000 signing bonus. Use competing offers from DeepMind, OpenAI, Anthropic, or CrowdStrike to push total comp 15–20% above initial offer. Accept-at floor: Cambridge £82,000 base / London £92,000 base / San Francisco $185,000 base — below these levels, the offer is below market for ML/AI engineers with cybersecurity expertise and you should counter aggressively on both RSUs and compute budget.

---

Evidence & Sources

• Darktrace "State of AI" Report — published Feb 3, 2026 — establishes the ML/AI research agenda for Agentic Identity defense and the Darktrace / SECURE AI product line
• Darktrace Investor Relations (LSE: DARK) — Annual Report, AI R&D investment, and model innovation metrics (FY2025)
• Levels.fyi — ML/AI Engineer compensation data for cybersecurity and AI companies (2025–2026)
• Glassdoor — Darktrace ML/AI Engineer salary reports, UK and US (2025–2026)
• AI Jobs Salary Survey — UK and US ML Engineer compensation benchmarks (2026)
• Cambridge AI Ecosystem Report — Local ML talent market and compensation data (2026)