Meta AI Support Assistant Archives

News
0
0
9 min read

Meta Expands AI Tools to Detect and Block Scam Messages

We are launching Meta AI Support Assistant worldwide on Facebook and Instagram. This AI-powered tool provides 24/7 automated assistance for account issues, such as real-time password updates and profile settings adjustments.

We plan to introduce progressively advanced AI systems across our apps to strengthen our enforcement capabilities. These systems are designed to quickly identify and remove serious content violations, such as scams and illegal material, by analyzing patterns, behaviors, and content for signs of misuse.

Today, we are launching new AI tools for support and content enforcement to improve your experience on our apps as technology advances. We are using AI to provide you with reliable help when you need it and to detect serious violations, such as scams, more quickly and accurately, while reducing pleading errors.

Introducing the Meta AI Support Assistant On Facebook And Instagram

In December, we previewed the Meta AI Support Assistant. This tool provides reliable 24/7 support for nearly any issue. Now we are rolling it out in areas where Meta AI is available on Facebook and Instagram for iOS, Android, and other desktop Help Centers. We’ve added even more capabilities and ways to help (updated March 20, 2026, at 1:30 p.m. PT)

When you have an issue, you need a solution – not just a suggestion. The new Meta AI support assistance is designed to help you resolve account problems from start to finish. It offers answers for any question – like about notification settings or new features – and issues like it can also take action for you on a growing set of requests directly within Facebook and, in the future, on Instagram, including:

Reporting scans, impersonation accounts, or problematic content

See why your account was removed, manage appeals, track updates, adjust privacy and profile settings, and reset your password.

Managing your privacy settings

Resetting passwords

Updating profile settings

Getting support should be simple. The Meta AI Support Assistant is built into Facebook and Instagram, so support is just a tap away, with responses in under five seconds—dramatically reducing wait times compared to traditional Help Center searches or seeking answers on external websites.

The Meta AI Support Assistant marks progress in our app support. Most feedback has been positive. It’s now available in all languages supported by Facebook and Instagram.

We started rolling out the Support Assistant to help people who have trouble logging into their Facebook and Instagram accounts. Right now, it’s available in select cases in the US and Canada. We plan to expand to more countries and support more account types and scenarios soon. We are continuing to invest in AI-powered tools to make support more accessible, reliable, and effective. As more people use the Meta AI Support Assistant and the technology improves, we will continue to refine it. Learn more about the Meta AI Support Assistant here.

Improving Content Enforcement With More Advanced AI

Last year, we shared positive results from changes we made to reduce mistakes and focus our proactive enforcement on the most serious illegal conditions: content such as terrorism, child exploitation, drugs, fraud, and scams. We also explained that we’re testing more advanced AI systems for content enforcement. These new systems can catch more violations accurately, stop more scams, and respond faster to real-world events, all while making fewer mistakes.

Early tests of these systems have been promising, promising as they can:

Increase detection rates for scam attempts that trick people into revealing login details. AI algorithms have helped identify and address up to 5,000 previously undetected scam attempts daily.

Identify and prevent additional accounts from impersonating celebrities and other high‑profile people, helping us reduce reports about the most impersonated celebrities by over 80%.

Double the detection of violent or explicit sexual solicitation using AI models, while also reducing missed violations by more than 60% through improved accuracy.

Prevent account takeovers by monitoring access patterns. If an account is accessed from an unusual location, or if the password or profile is changed unexpectedly, AI flags this behavior potentially harmful.

Detect fake websites imitating legitimate ones by analyzing suspicious URLs, logos, and pricing patterns. After broader testing, AI helped reduce user exposure to scam ads and other severe violations by 7%, providing stronger user and brand protection.

These advanced AI systems can do all of this in languages spoken by 98% of people online. This is much more than the previous coverage of about 80 languages. They can also increase capacity in any language as needed. They adapt to understand cultural traits, including niche subcultures, rapidly changing and regionally specific code words, emoji meanings, and slang.

A Smarter Approach

In the next few years, we’ll start using these advanced AI systems across our apps once they consistently outperform our current content enforcement methods. This will change our approach as we make this shift. We will rely less on third-party vendors for content enforcement and focus on strengthening our own systems and teams. We still have people reviewing content, but these AI systems will handle tasks better suited to technology, such as repetitive reviews of graphic content or areas where bad actors are constantly changing their tactics, like illicit drug sales or scams.

Even as we use new technology to expand what we can do, people will stay at the center of our approach. AI helps us move faster and work at a larger scale, but it does not replace human decision-making. Instead, it helps us apply that judgment more consistently across billions of pieces of content. Experts will design, train, oversee, and evaluate our AI systems, measure their performance, and make the most complex and important decisions. For example, people will still play a key role in making the highest-risk and most critical decisions, such as appeals of account disablement or reports to law enforcement.

We thoroughly test each AI system, install safeguards, and evaluate performance to minimize bias and ensure consistency and accuracy. With AI-driven tools like the Meta AI Support Assistant, people can more easily report violations and appeal errors, letting AI handle scalable detection while experts manage nuanced decisions. This combined approach leverages both advanced AI and human expertise for optimal support and safety.

Source: Boosting Your Support and Safety on Meta’s Apps With AI

News
0
0
6 min read

Meta Launches AI Tools to Help Users Recover Hacked Accounts

On March 25, 2026, Meta announced new AI tools to help users recover hacked or compromised accounts, replacing the confusing and slow manual support systems that Facebook and Instagram users have struggled with for years.

For security experts and social media strategists, this update is more than just convenient. It changes identity verification, shifting from hardware checks to smarter AI-driven processes by adding advanced AI meta, aiming to scale support and better protect accounts.

Introducing the Meta AI Support Assistant

The main feature of this update is that the Meta AI Support Assistant is now available around the clock in Facebook and Instagram help centers. Unlike older chatbots that relied on keywords, this assistant uses Meta’s latest Llama technology to comprehend and respond to natural language questions with much better context.

Now, if a user gets locked out, they don’t have to search through endless help articles. They can talk directly to the assistant, who can perform tasks such as resetting passwords or updating profiles right in the app. Meta says the assistant usually responds in less than 5 seconds, making a big difference for users during stressful security issues.

Video Selfies: Biometric Verification at Scale

In March 2026, a major change is introduced: video selfie verification instead of uploading photos of government IDs, which can be easily faked with deepfake tech. Users now use this more secure method.

The video selfie process is like today’s smartphone face recognition. Users record a short video while moving their head, which allows the system to create a 3D face map. Meta’s AI checks this video in real time against the profile and past media images.

This feature, called liveness detection, sets the system apart by asking users to make random movements. It can tell the difference between a real person and a photo or deepfake video for business users or public figures. This provides a level of security that regular two-factor authentication via SMS, which can be compromised through SIM swapping, does not offer.

Adaptive Recovery And Trusted Device Authentication

The update also introduces adaptive recovery flows, in which the AI security system evaluates multiple data points to assess the risk of account recovery attempts. The system evaluates factors such as login location, account activity patterns, device fingerprints, and past access history to make informed decisions about granting or denying access.

If someone recovers their account from a familiar device and place, the AI speeds up the process. If it detects several suspicious changes, such as a new password, email address, or location, it switches to stricter checks. Already, this has reduced account takeovers by 30% compared to 2025.

Privacy Safeguards and Data Governance

Given Meta’s history with facial recognition, its handling of new biometric data is under scrutiny. Meta states that video selfies serve only for identity verification and are not visible to others.

To help protect privacy, Meta uses a zero-retention policy for biometric data from these sessions. After the AI verifies a user’s identity, the facial information is immediately deleted from the servers. This prevents the creation of a large database of face prints that could be at risk in a data breach. In places with strict privacy laws, such as the European Union and some U.S. states, users can choose a manual review by a person instead of using the AIR system.

Combatting the Celeb Bait Scourge

The AI also targets celeb-bait ads and fake promotions that use celebrities’ images to trick users. The facial recognition system scans ads on the platform for these scams.

If the system detects a public figure’s image in an ad, it compares the face to the public figure’s official profiles. Fake matches lead to immediate ad blocking. Automated checks are essential in 2026, as AI-generated scam ads now exceed the capacity of human reviewers.

The Sentinels of the Digital Threshold

These AI tools create a stronger, more reliable digital system. It’s like our accounts have an immune system that tells real from fake. Soon, forgetting a password could just mean a quick check between you and the machine that recognizes you. The future of account security may lie in the uniqueness of your face, keeping your accounts safe from intruders.

Source: Meta AI support assistant

News
0
0
7 min read

WhatsApp Adds AI Tools to Detect and Block Scams

Meta AI Support Assistant now offers 24/7 help on Facebook and Instagram for account updates and settings.

In the coming years, we will introduce advanced AI systems across our apps. These will help us find and remove serious content violations, such as scams and illegal material, so you encounter them less.

We are launching new AI tools to support and enforce content on our apps, enhancing your experience. As technology advances, we are using AI to provide more reliable assistance and detect serious violations, like scams, faster and more accurately, with fewer mistakes.

Meta AI Support Assistant is now available on Facebook and Instagram

In December, we introduced the Meta AI Support Assistant, a tool built to offer reliable 24/7 support for almost any issue. Now it’s available in countries and regions where Meta AI is supported on Facebook and Instagram for iOS and Android, as well as on the desktop Help Center, with even more features to assist you. Updated on March 20, 2026, at 1:30 PM PT to reflect availability.

When you have an account Issue, you want solutions, not just answers. The Meta AI Support Assistant is designed to resolve problems from start to finish. It can answer questions about notifications or new features, and handle follow-up requests in Facebook and Instagram, such as:

Reporting scams, impersonation accounts, or problematic content.

Making it easier to see why your account was taken down, to appeal, and to track what happens next.

Managing your privacy settings.

Resetting passwords.

Updating profile settings.

The Meta AI Support Assistant is built into Facebook and Instagram, responding in under 5 seconds to reduce wait times. Meta AI Support Assistant improves app support. Most feedback is positive. It’s available for all support topics in supported languages.e.

We’ve begun offering support to people who need help logging into Facebook and Instagram, starting with certain cases in the US and Canada. We plan to expand to more countries and account access scenarios soon. We are investing in AI-powered tools to make support more accessible, reliable, and effective. As more people use Meta AI support assistance and as technology improves, we’ll keep improving it. Learn more about the Meta AI support assistant here.

Improving Content Enforcement With More Advanced AI

Last year, we shared positive results from changes designed to reduce mistakes and focus our enforcement on the most serious content, such as terrorism, child exploitation, drugs, fraud, and scams. We’ve also been testing more advanced AI systems for content enforcement. We think these systems can catch more violations accurately, stop more scams, and respond faster to real-time events, all with fewer mistakes.

Early tests of these systems have been promising, as they can:

AI identifies 5,000 scam attempts daily that review teams missed, reducing login detail theft.

Identify and prevent additional accounts from impersonating celebrities and other high-profile figures, helping us increase user reports about the most impersonated celebrities by over 80%.

Catch two times more violating adult sexual solicitation content than our revisions, while also decreasing the rate of mistakes by more than 60%.

AI stops account takeovers by flagging new locations or profile edits as threats.s.

Advanced AI can spot scam sites by recognizing fake web addresses, irregular prices, and misuse of real logos, reducing scam ad views by 7%.

These AI systems cover languages used by 98% of people online, adapting to different cultural and regional traits.

A Smarter Approach

In the next few years, we will deploy these advanced AI systems across our apps once they consistently surpass our current content enforcement methods. As we do this, we will rely less on third-party vendors and focus on building up our own systems and teams. While people will still review content, these AI systems will handle tasks better suited to technology, such as repetitive reviews of graphic content or areas where bad actors often change tactics, such as illicit drug sales or scams.

Even as we use new technology to expand what’s possible, people will remain central to our approach. AI lets us move faster and scale, but doesn’t replace human judgment. Instead, it applies that judgment more reliably to billions of pieces of content. Experts will design, train, oversee, evaluate, and measure our AI’s performance and make the most complex decisions. For example, people will still handle the highest-risk and most critical issues, such as account-disability appeals. Experts will design, train, oversee, evaluate, and measure our AI’s performance and make the most complex decisions. Once to prevent bias and ensure accuracy. Our community standards remain unchanged, with tools like the Meta-AI support assistant. We’re improving how people report policy violations and appeal mistakes. Ultimately, this approach lets people focus on their strengths while technology handles what it does best, merging advanced AI with human expertise for mutual support.

Source – Boosting Your Support and Safety on Meta’s Apps With AI