AI
Content
Accessibility

AI turns TV archives into digital gold at ANT1 TV

A frame of an old series

For years, like most public broadcasters, Antenna Group sat on a treasure trove of 90s television. The problem? Most of it was locked away in aging formats, with fuzzy images, noisy sound and almost no metadata. Great for nostalgia, but terrible for streaming. So, the Greek media group decided to do something bold: let AI loose on its past. The result is a full-scale archive revival that is turning dusty shelves into one of the hottest content sources on its ANT1+ platform. 

From forgotten tapes to flagship content 

Under the leadership of Antenna Group CTO Costas Colombus, an ambitious AI-project was launched around one of its biggest challenges: 7,500 episodes from the 1990s, many without proper metadata or up-to-date technical standards. Instead of treating the archive as a cost center, the team treated it as a product. The mission was simple: make old content look good, sound good and be easy to find. 

To do that, Antenna stacked several AI technologies on top of each other: 

  • Super resolution tools upgraded low resolution footage to HD and even 4K.
  • Automated denoising cleaned up decades-old audio tracks when needed, creating even audio stems for any further use.
  • Speech to text systems generated transcripts with around 95% accuracy.
  • AI models then used those transcripts to create synopses and metadata for the ANT1+ streaming platform. 

In other words, they did not just restore shows. They made them searchable, browsable and bingeable. 

The numbers that changed the business case 

The impact was immediate and very tangible. Some of the enhanced archive shows started climbing into the top three most viral titles on ANT1+. Content that had been practically invisible suddenly became prime streaming material. At the same time, AI generated metadata made it easy for viewers to discover these series in the first place, while automatically generated thumbnails boosted click-through rates. 

For ANT1 TV, the archive stopped being a line item in the budget and started behaving like a new revenue stream. 

Inside the AI toolbox 

To pull this off, Antenna Group relies on a mix of specialized tools: 

  • ACRCloud, which recognizes songs and music cues in seconds so rights and clearances can be managed efficiently
  • Audioshake, which splits mixed audio into separate stems so dialogue, music and effects can be cleaned up or redubbed
  • TOPAZ Labs video enhancement tools
  • Speechmatics Speech to text technology 

On top of that, the company has begun experimenting with AI avatars in niche content verticals like Formula 1, astrology and training videos. It is a clear signal that for Antenna, AI is not just about fixing pixels. It is about inventing new kinds of programming. 

Looking ahead to 2026 

The next phase is all about personalization and automation at scale. ANT1 TV is exploring: 

  • Personalized content recommendations
  • Smarter ad optimization and performance tracking
  • Automated trailer and promo creation 

By 2026, the company expects the economics of localization to be dramatically different, as AI further reshapes dubbing and subtitling workflows across the industry. 

What other media companies can learn 

ANT1 TV’s story is a reminder that AI projects succeed when they start from business questions, not from shiny tools. Here, the central question was: how do we turn an expensive archive into a profitable, living catalogue? 

A few key lessons emerge: 

  1. Start with a clear use case 
    Focus on specific, measurable outcomes, such as transcription accuracy, per episode cost reductions or engagement metrics on streaming platforms.
  2. Scale gradually 
    Begin with one archive segment or series, prove the value, then expand AI use as internal expertise and tools mature.
  3. Think beyond restoration 
    AI can clean up footage, but its real power shows up when you use it to rethink formats. From auto generated trailers or short clips and data driven recommendations, all build on the same foundation as restoration. 

In ANT1 TV’s, this mindset turned archives into what the company now treats as a living resource that can speak to new generations of viewers, not just a museum of past programming. 

ANT1 TV’s archive now breathes again: digitally, intelligently and with purpose. With AI acting as both engineer and creative partner, the past is no longer merely preserved. It is reimagined for the future. 

Article written by Kobe Briers with the input of Costas Colombus (Antenna Group).