Elfbot moderates your comment sections around the clock. It hides hate and disinformation before your team sees it, answers genuine questions with verified facts, and thanks the people who stand with you.
Every comment,handled.
Elfbot reads each comment the moment it lands and sorts it into one of four kinds. Each kind gets the response it deserves.
Hidden before anyone sees it.
Threats, abuse and coordinated disinformation are hidden within seconds, following your moderation rules. Your staff never have to read them to do their job. Like trolls in the old stories, they turn to stone in daylight.
- Hidden, not deleted. Every comment is archived.
- Severity scored, with a plain reason for each decision.
- Unhide any comment with one click.
Brussels traitors. Someone should visit your office and teach you a lesson.
Brussels traitors. Someone should visit your office and teach you a lesson.
Answered with facts, in your voice.
When people ask real questions, or repeat a false claim they’ve picked up, Elfbot checks your organisation’s own facts and trusted sources, then drafts a reply with the evidence attached.
- Fact-checks against news and official sources.
- Replies in the commenter’s language.
- You choose which replies need approval.
Is it true the EU is banning cash from 2027?
Fact-check
The EU is banning cash in 2027✕ FalseNews and official sources
There is no ban. From July 2027 an EU-wide €10,000 limit applies to cash payments to businesses. Cash stays legal tender.
Suggested reply · en
Love sent back, every time.
Your supporters deserve an answer too, but your team is busy. Elfbot learns your tone of voice and thanks the people who back your work, so nobody who shows up for you is left unheard.
Thank you for keeping this a space where we can actually talk. Your work on fair elections matters.
Flagged for escalation or DSA reporting.
Content that may break the law, like incitement or illegal hate speech, is archived with its evidence and prepared as a notice to the platform under the Digital Services Act. Your team signs off.
- Comment, author and URL archived with a timestamp
- Classified as possible incitement to hatred
- DSA Article 16 notice drafted for the platform
- Your escalation lead notified
Hide the hate.Keep the debate.
Public institutions can’t be seen silencing critics, and Elfbot won’t make you. Disagreement stays visible. Only abuse, threats and coordinated manipulation are hidden.
Get lost. You’re all idiots!HiddenHidden according to your moderation rules.
I disagree with your migration policy.VisibleStays up. Respond using your guidelines.
You voted against the ceasefire resolution.CheckedEvidence checked, reply drafted with sources.
Thank you for speaking up.ThankedAcknowledged, so you don’t have to every time.
Built for the institutionsdisinformation targets.
Election bodies, EU and UN agencies, embassies, ministries and the NGOs around them are prime targets for coordinated hate and false narratives. Elfbot is made for that reality.
Protects democracy.
Coordinated comment disinfo is designed to create the illusion of consensus – the gov is corrupt, the UN is biased – because more often than not, people follow the majority. Elfbot stops the trolls from achieving their objective.
Protects staff.
Reading threats and slurs all day wears people down. With Elfbot, your communications team never has to be exposed to hate and lies again. Only borderline cases needing human sign off get sent to them.
Protects the organisation.
Narrative attacks are hidden in real time. Every decision is logged with a reason, every comment archived, and nothing happens outside of the approval settings you choose.
Win back time for the real work.
Constituents, campaigns, the mission. Put in your own numbers.
hours a week spent reading, sorting and answering comments.
That’s about 123 working days a year.
Your estimate, based on the numbers you entered and 7.5-hour days over 46 working weeks.Guidelines
Your voice.Your rules.
Write your moderation policy in plain language. Who you are, what supporters sound like, where criticism ends and abuse begins. Elfbot follows it in every decision and every reply, with whatever avatar you choose.
Connect. Set your rules.Let Elfbot work.
- 01
Connect your page
Link the Meta account you want to manage. It takes four clicks and no training.
- 02
Add your facts
Give Elfbot your key information, your positions and your preferred tone.
- 03
Choose automation
Preview replies and decide what runs on its own and what needs approval. Monitor as much as you like.
Let Elfbot takethe next shift.
Included in your plan
- AI-assisted moderation and reply drafts
- Fact-check credits every month
- Your rules and approval settings
- Escalation and DSA reporting flags
Each month: 10,000 comments, up to 1,000 replies and 50 fact-check credits. Running pages across several delegations, country offices or agencies? Write to us about a multi-page setup.
Questions
Will Elfbot hide criticism of our organisation?
No. Your policy sets the line between criticism and abuse, and disagreement stays visible by default. Elfbot hides abuse, threats and coordinated manipulation, and gives a reason for every decision so you can check it.
Are hidden comments deleted?
No. Hidden comments stay on the platform, hidden from public view, and every comment is archived in Elfbot. You can unhide any of them with one click.
Does Elfbot post without a human checking?
Only where you allow it. You choose which kinds of reply post automatically, such as thanking supporters, and which wait for a person to approve them. Escalations and DSA notices always wait for your sign-off.
What does DSA reporting mean here?
The EU Digital Services Act lets anyone notify a platform about content they believe is illegal. When Elfbot flags something as possibly illegal, it archives the evidence and drafts that notice for your team, ready to review and send, or to pass to your escalation lead.
Which platforms and languages does it cover?
Elfbot connects to the pages on your Meta account. You set your community languages in your guidelines, and Elfbot replies in the language the commenter used.
Do we need training to use it?
No. It connects in four clicks and the settings are written in plain language. Our support team is there if you want help setting up your policy.