What Security+ 801's New AI Objectives Mean for Your Study Plan

At a Glance

SY0-801 launches November 17, 2026, and AI is now on the test. CompTIA's published objectives add one dedicated AI objective (2.6) with 13 listed topics, plus AI line items inside three other objectives: LLMs as an attack surface (2.4), deepfakes (2.5), and AI capabilities in automation (4.6). Budget roughly a week of focused study for it. Not a month.

Cybersecurity and IT

CompTIA has published the full SY0-801 objectives. No guessing required. Here's what's actually in the document, where it sits, and how to fold it into a study plan you may already be running.

What has CompTIA actually published about SY0-801?

Everything you need to plan. The Security+ V8 page lists the exam details, and it links to the full objectives document (Version 2.0).

	SY0-801 (V8)
Launch date	November 17, 2026
Questions	Maximum of 90, multiple-choice and performance-based
Duration	90 minutes
Passing score	750 (scale of 100 to 900)
Recommended experience	Security administrator with two years of hands-on experience
Retirement	Estimated 3 years after launch

Domain weights: General Security Concepts 16%, Threats, Vulnerabilities, and Attacks 24%, Security Architecture 19%, Security Operations 27%, Security Program Management and Oversight 14%.

SY0-701 is still available. It retires June 11, 2027 for English. If you're deciding between versions, start with our breakdown of what's changing in SY0-801.

Where does AI show up in the SY0-801 objectives?

Four places. One is a full objective. Three are line items inside existing objectives.

Objective	Verb level	AI content listed
2.4 Explain types of vulnerabilities and attack surfaces	Explain	Large language models (LLMs) as an attack surface
2.5 Given a scenario, analyze indicators of malicious activity	Scenario	Deepfake, under social engineering attacks
2.6 Summarize threats and vulnerabilities associated with artificial intelligence (AI) usage	Summarize	Model manipulation, poisoning, prompt injection, data loss, bias, explainability, hallucinations, jailbreaking, evasion, privacy, ethical considerations, session hijacking, code execution
4.6 Given a scenario, apply automation and orchestration solutions to secure operations	Scenario	AI capabilities: agentic, chatbot, predictive analysis, AI-augmented baselines. Guardrails listed under considerations

AI and LLM also appear on the official acronym list.

That's real coverage. It's also not a rewrite. The other objectives in Domain 2 and the rest of Domain 4 still carry most of the weight in those domains.

Which AI topics deserve the most time?

Read the verbs. CompTIA tells you how deep each objective goes.

2.6 says "Summarize." That's a recall-and-recognize level. You need to define each of the 13 terms, tell them apart, and match each to an example. You don't need to build or attack a model.

4.6 says "Given a scenario, apply." That's deeper. Expect questions that describe a SOC workflow and ask what automation or AI capability fits, or what guardrail is missing. This is where AI can show up in a performance-based format.

2.5 says "Given a scenario, analyze." Deepfake sits next to phishing, vishing, and impersonation. Expect it inside a social engineering scenario, not as a standalone definition.

So: memorize 2.6. Practice 4.6. Recognize deepfakes in context.

The 13 terms in objective 2.6, in plain English

Learn these as pairs and contrasts. That's how the exam separates them.

Term	What to know
Prompt injection	Input crafted to override the model's instructions. Can arrive directly or hidden in content the model reads.
Jailbreaking	A prompt technique to get past a model's safety restrictions. Related to prompt injection, not identical.
Poisoning	Tampering with training or fine-tuning data so the model learns the wrong thing.
Model manipulation	Altering a model's behavior or weights after training.
Evasion	Crafting input so a model misclassifies it at inference time. Think malware built to slip past an ML detector.
Data loss	Sensitive data leaving through prompts, outputs, or logs.
Privacy	Personal data exposure through training data, prompts, or responses.
Hallucinations	Confident output that's false.
Bias	Systematically skewed output from skewed data or design.
Explainability	Whether you can tell why the model produced an output. Matters for audit and incident response.
Ethical considerations	Fairness, accountability, and appropriate use.
Session hijacking	Taking over an authenticated AI session or its tokens.
Code execution	A model or agent running code, or its output being run, in a way an attacker can steer.

Want a second lens? The OWASP Top 10 for LLM Applications (2025) covers several of the same risks from a builder's view: prompt injection (LLM01), sensitive information disclosure (LLM02), data and model poisoning (LLM04), and misinformation (LLM09). That mapping is our study aid. It isn't CompTIA's. Use it to understand the concepts, then answer in CompTIA's terms.

How to fit AI into your study plan

Two rules. Don't bolt it on at the end. Don't let it eat the schedule.

If you're starting fresh on SY0-801, reweight your plan to the V8 numbers. Security Operations is still the biggest domain at 27%. Domain 2 grows to 24%. Domain 5 drops to 14%. Then slot the AI work into the domains where it lives:

Week block	Domain focus	AI work to add
Domain 2 weeks	Threats, Vulnerabilities, and Attacks (24%)	Learn all 13 terms in 2.6. Add LLMs to your attack-surface notes for 2.4. Add deepfakes to your social engineering flashcards for 2.5.
Domain 4 weeks	Security Operations (27%)	For 4.6, compare agentic, chatbot, predictive analysis, and AI-augmented baselines. Write one example use case and one guardrail for each.
Final review week	All domains	Run a timed practice set. Tag every AI miss by objective number.

That adds up to roughly a week of AI-specific effort across a typical plan. It's a planning estimate, not a CompTIA figure. Adjust for your background.

If you're mid-plan on SY0-701, you don't need any of this for your exam. Finish. SY0-701 doesn't carry a dedicated AI objective. Our 8-week SY0-701 study plan still applies through June 11, 2027.

If you're on the fence, use the AI content as a tiebreaker. Already using AI tools at work? The 801 material will feel familiar. Starting from zero on AI? Factor in that extra week.

How do you practice AI objectives without an AI lab?

Hands-on beats flashcards here, especially for 4.6.

Use an approved AI assistant on a safe task. Summarize a public advisory. Then check every claim against the source. You just practiced spotting hallucinations.
Write a guardrail. Pick an automation, like auto-closing low-severity tickets. Write down what the AI is allowed to do, what needs human approval, and what gets logged. That's 4.6 in practice.
Read a prompt injection example. Read OWASP's description of LLM01, then explain in two sentences why a SIEM that feeds log text to an AI assistant could be exposed. That's the analysis level the exam rewards.

For a structured version, AI Essentials in the ACI ITPro library builds the vocabulary and the habits with hands-on ACI Skill Labs. More on picking an AI path: AI Essentials vs. SecAI+.

Build Your Security+ Foundation

Get Security+ courses, hands-on ACI Skill Labs, and practice exams in one 12-month plan, so you walk into exam day ready.

Get the Security+ Bundle

Not sure Security+ is your first step?

Browse every CompTIA certification bundle to find the right starting point. If you want AI Essentials and the rest of the AI courses with hands-on labs, compare ACI ITPro library plans. Still deciding how deep to go on AI? Read AI Essentials vs. SecAI+.

November 17, 2026, per CompTIA's Security+ V8 page.

Yes. Objective 2.6: "Summarize threats and vulnerabilities associated with artificial intelligence (AI) usage." It lists 13 topics, including prompt injection, poisoning, hallucinations, and jailbreaking.

Objective 2.4 lists LLMs as an attack surface. Objective 2.5 lists deepfakes under social engineering. Objective 4.6 lists AI capabilities (agentic, chatbot, predictive analysis, AI-augmented baselines) under automation and orchestration.

SY0-701 has no dedicated AI objective. If you're testing on SY0-701 before it retires June 11, 2027, you don't need to add the 801 AI material.

Roughly a week of focused work across a typical plan is a reasonable estimate. It's a study-planning convention, not a CompTIA figure.

No. Security+ covers AI at a foundational level. CompTIA SecAI+ (CY0-001) is a separate, advanced certification for people who secure AI systems as part of the job.